Teleperformance

Reference: pages 82-83

Sustainability governance sits with the Board of Directors, which had 13 members at 31 December 2025, "including 9 independent members, representing an independence rate of 69% according to the definition in ESRS 2 Gov 1 standards", and "an average female representation rate of 48% during the year 2025" (p.82). Two employee representative directors sit on the Board (Veronique de Jocas and Evangelos Papadopoulos), and the full composition table with gender, independence and executive status is printed on p.83.

Two Board committees share the work. The Sustainability Committee (three members, meeting at least three times a year) "verifies the integration of the Group's social and environmental commitments, reviews regulatory publications and assesses impacts, risks and opportunities in consultation with the Audit, Risk and Compliance Committee", and reviews the sustainability report and the Vigilance Plan (p.83). The Audit, Risk and Compliance Committee "is responsible for monitoring the preparation and integrity of sustainability reporting, as well as monitoring internal control and risk management systems".

Expertise is described rather than tabulated: the Sustainability Committee is chaired by an independent director who is "an experienced human resources manager in a large Colombian group", and an employee representative director who is a Global Social Auditor also sits on it (p.83).

Reference: page 83

"The Board of Directors is kept informed of impacts, risks and opportunities, as well as the implementation of due diligence, through regular reports presented by its Sustainability Committee at Board meetings" (p.83). The Committee reports to the Board and meets at least three times a year.

The report names the sustainability matters actually taken during the reporting period. "During the reporting period, the following topics were addressed: implementation of the CSRD; environmental roadmap; TP's approach to human rights; sustainability reporting and non-financial ratings; the non-financial criteria applicable to executive remuneration" (p.83).

The Audit, Risk and Compliance Committee's work "includes verifying sustainability reporting, as well as the internal control and risk system, in conjunction with the Sustainability Committee and sustainability auditors". Other information routes are set out in the stakeholder table on p.86, which records for each stakeholder group the "Method of informing directors and executive officers": summaries of discussions with staff representatives presented at Board meetings, supplier assessment results presented to the Sustainability Committee, ESG ratings presented to management and the Board, and compliance reports and inspection summaries forwarded to the Board.

Not disclosed: how the Board addressed trade-offs between material IROs, or which decisions during the year turned on sustainability information.

Reference: pages 84-85

Both the annual and the long-term incentive carry sustainability conditions, and the 2025 outcomes are disclosed.

Annual variable remuneration 2025. "subject to financial performance conditions (accounting for 60%) and non-financial performance conditions (accounting for 40%)" (p.84). Two of the non-financial criteria are described in full:

• Employee engagement, 6.66 points, split 50/50 between the share of employees in a certified subsidiary (target 90%; "TP has obtained or renewed certifications in 69 countries. These certifications cover 90% of the Group's workforce") and the Trust Index (target above 70%; achieved 77%). The Board "noted the full achievement" of both sub-criteria.
• Cybersecurity, 6.68 points - completion of the second phase of Project Eagle and progress on the third, all ISO 27001, ISO 27701, PCI and HIPAA assessments completed, "more than 580 clients audits with no open findings". Fully achieved (p.85).

Long-term incentive. The performance share plan of 31 July 2025 (Plan 250731TP), 769 beneficiaries and 760,075 shares, is measured over 2025-2027 on 75% financial and 25% extra-financial criteria: a "CSR" environment criterion at 12.5%, based on Scope 1 and 2 reduction aligned with the 2030 targets (100% vesting at a reduction of 41.3% or more), and a "Promotions" criterion at 12.5% on the internal promotion rate (100% at 60% or more).

Reference: page 84

The due diligence statement is given as a mapping table rather than a narrative. "TP implements a due diligence process to identify, prevent and reduce its environmental, social and governance impacts", and the five core elements are cross-referenced to the sections that carry them (p.84):

• Embedding due diligence in governance, strategy and business model - section 3.1.3
• Engaging with affected stakeholders in all key steps of the process - section 3.1.4
• Identifying and assessing adverse impacts - section 3.2.1
• Taking action to address those adverse impacts - sections 3.3 to 3.5
• Tracking the effectiveness of these efforts and communicating progress - sections 3.3 to 3.5

The IRO-2 Appendix B cross-reference table confirms the placement, listing "ESRS 2 GOV-4 Statement on due diligence, paragraph 30" against section 3.1.3 (p.139).

Substantive due diligence machinery is described elsewhere and is unusually concrete for this DR: the CHRB-based self-assessment of human rights commitments and the annual reassessment against the latest CHRB standard (p.114), the network of local human rights experts feeding local risk maps into the Group mapping (p.115), and the supplier funnel with 596 high-gross-risk suppliers identified in 2025 (p.128). Governance of the Vigilance Plan is jointly coordinated by the Sustainability Department, Human Resources and the Compliance, Privacy and Security teams (p.110).

Reference: pages 83, 135

The internal control disclosure is short. "TP has incorporated sustainability issues into its internal control questionnaire (in particular on the existence of a reference framework and the validation of data by local CFOs), and is making every effort to develop operational controls in its subsidiaries" (p.83). Oversight sits with the Audit, Risk and Compliance Committee, which "is responsible for monitoring the preparation and integrity of sustainability reporting, as well as monitoring internal control and risk management systems" and reviews the internal control and risk system "in conjunction with the Sustainability Committee and sustainability auditors" (p.83).

The data controls are set out in the methodology note (p.135): environmental data is collected three times a year and "checked by the Chief Financial Officers of each subsidiary and, at Group level, by the Sustainability Department", staff data is monitored by the Reporting and Consolidation Department via consistency checks, and qualitative data covers 94% of the workforce.

Not disclosed: the scope and frequency of the risk assessment over sustainability reporting, the risks identified, or how internal control findings are reported to the administrative bodies. The management-report cross-reference table points GOV-5 to page 90 (p.351), which is the double materiality methodology rather than a control description.

Reference: pages 76-81

TP describes itself as a provider of "digital integrated business services", managing daily interactions with consumers on behalf of clients. Scale figures for 2025 (pp.78-80): revenue of EUR 10,209 million, EBITA before non-recurring items of 14.6%, net free cash flow of EUR 901 million, nearly 490,000 employees, over 600 facilities, around 170 markets, more than 400 languages and dialects, over 1,500 clients and around 4,000 suppliers. Environmental resources are given alongside financial ones: 551,700 tCO2e emitted across Scopes 1, 2 and 3 and 411,110 MWh of electricity used.

The value chain is set out on pp.80-81. Upstream: around 4,000 suppliers of "Labor services; Security, cleaning; Advisory services; Employee transport; Office rental; IT equipment; Software, telecoms", plus infrastructure (over 600 facilities and the TP Cloud Campus work-at-home solution) and energy "related to office use and digital footprint (cloud, mostly outsourced data centers)". Downstream: over 1,500 clients, consumers and citizens, and "Communities and the planet".

Excluded activities are stated plainly, which is the ESRS 2 paragraph 40(d) datapoint: "the Group does not offer any prohibited products or services in the markets in which it operates. The Group does not operate in sectors related to fossil fuels, chemical production, controversial weapons or tobacco" (p.76). The Appendix B table maps those four datapoints to sections 3.3.3.3 and 3.1.1 (p.139).

Reference: page 86

Stakeholder engagement is presented as a single table covering eleven groups: TP employees, staff representatives, non-employees specific to LanguageLine Solutions, workers in the value chain, clients, end-users, investors, rating agencies, local communities, government authorities and regulators, and suppliers. Each row records the stakeholders' interests and viewpoints, the organizational procedures used, the purpose of the dialog, the follow-up of conclusions, the changes made to strategy or business model, and the method of informing directors and executive officers (p.86).

Worked examples from the table:

• Staff representatives - interests in wellbeing at work, health and safety and social dialog; "Quarterly and/or annual meetings depending on local practice"; outcome "Establishment of collective agreements"; results presented to the Board of Directors via staff representatives.
• Workers in the value chain - interests in "Respect for fundamental rights; Working conditions"; engagement through "Compliance with the Code of Conduct, supplier assessment"; outcome "Corrective measures with regard to suppliers, support"; assessment results presented to the Sustainability Committee.
• End-users - interests in "Simple and rapid solutions to their daily requests; Privacy"; satisfaction surveys and omnichannel contacts; monitoring of satisfaction rate and service quality.

Reference: pages 90-95

The materiality matrix on p.90 plots each topic on impact materiality against financial materiality, with the rule stated: "An IRO is considered material if its score exceeds 2.5 out of 5, a threshold chosen to reflect a significant impact on stakeholders, the environment, or the company's economic performance."

Material IROs are then described in tables spanning pp.91-95, grouped under the standard each belongs to. Counting the rows gives 29 individually typed IROs: ESRS E1 greenhouse gas emission reduction and climate change adaptation (3), ESRS E5 e-waste management (2), ESRS S1 working conditions, equal treatment, labor relations and social dialog, health and safety, career development and employee data privacy (17), ESRS S2 human rights in the value chain (1), ESRS S4 consumer and end-user data security (2) and ESRS G governance and ethics and anti-corruption (4).

Each row carries a type of effect, a scope (operations, upstream, downstream or entire value chain), the impact or financial effect, the management response, a criticality rating and a time horizon.

Topics assessed and found not material are listed explicitly (p.91): responsible sourcing of materials, biodiversity, pollution and water; fundamental rights of end-users, protection of vulnerable users and impact on local communities; political commitment and influence, and supplier relations.

Reference: pages 89-90

The double materiality analysis "is coordinated by the Sustainability Department", worked with in-house experts, regional teams and staff representatives, and "is approved by the Sustainability Committee, the Audit, Risk and Compliance Committee and senior management" (p.89). The Steering Committee translated each ESG matter into IROs with the legal and compliance, finance, audit and internal control, human resources, ethics, health and safety and procurement functions.

Inputs are itemised: previous materiality analyses, the Group ERM risk map, vigilance-plan thematic risk mapping on corruption, human rights, environment, health and safety and suppliers, and employee, client and end-customer satisfaction surveys; externally, ISO 26000, the SDGs, GRI, the ESRS, industry benchmarks, media watch, and "international indexes to assess the level of country-specific gross risk affecting operations and the value chain (Global Climate Risk Index, Human Rights Index Score, Corruption Perceptions Index)".

Scoring is spelled out on p.90: impact materiality multiplies scale, extent and irremediable character, then likelihood, divided by five; financial materiality multiplies severity by probability, divided by five, with the threshold set in line with the ERM.

For 2025 the Group "continued to refine its assessment"; the auditor records that the entity determined "no significant changes were made to the double materiality assessment" (p.142).

Reference: pages 135, 139-140

The report does not print a full ESRS content index. Section 3.6 states: "The CSRD and GRI cross-reference tables, which provide an overview of the important sustainability information contained in the Universal Registration Document and other public documentation, may be found on the Group website" (p.135). A reader of the document therefore has no disclosure-requirement-by-disclosure-requirement concordance.

What is printed under the IRO-2 heading is narrower. Section 3.6.3, "IRO-2 Appendix B cross-reference table" (pp.139-140), lists the datapoints derived from other EU legislation and maps each to a section of the report. It covers rows for ESRS 2 GOV-1, GOV-4 and SBM-1, E1-1, E1-4, E1-5, E1-6, E1-7 and E1-9, E5-5, S1-1, S1-3, S1-14, S1-16 and S1-17, S2-1 and S2-4, S3-1 and S3-4, S4-1 and S4-4, and G1-1 and G1-4, and marks the E2-4, E3-1, E3-4 and E4-2 rows "Non-material".

A second, coarser table appears outside the sustainability statement. The cross-reference table to the management report (pp.351-352) lists ESRS 2 BP-1, BP-2, GOV-1 to GOV-5, SBM-1 to SBM-3, IRO-1, IRO-2 and MDR-P, MDR-A, MDR-M and MDR-T against page numbers, but handles the topical standards at standard level only (ESRS E1 to E5, S1 to S4 and G1), with ESRS E3 shown as "n/a".

The practical consequence is that coverage of the topical disclosure requirements has to be inferred from section headings, because the concordance is published elsewhere.

Reference: pages 97-102

The plan exists and is described, but its approval status is stated plainly and is a finding in itself: "The environmental risk mapping exercise helped to establish the Group's climate action roadmap. The transition plan was presented to the Board of Directors' Sustainability Committee and is currently pending approval" (p.100).

Content of the plan (p.100). "The strategy adopted to mitigate risks focuses on: increasing the proportion of renewable energy in electricity consumption to reach at least 50% by 2026 and 80% by 2030; achieving high energy performance at the Group's facilities by adopting efficiency measures; streamlining IT infrastructure by adopting measures to reduce energy consumption and increase the share of renewable energy in data centers and by purchasing STAR-rated and EPEAT-certified computer equipment; sensitizing its ecosystem (suppliers, customers, employees) to sustainable practices."

Dependencies are acknowledged: delivery "depends, among other things, on the availability of certified renewable energy sources in the countries in which the Group operates, as well as the ability to develop energy efficiency measures alongside lessors for the facilities, the majority of which are leased".

Alignment is asserted with a benchmark test: "TP is included in the Paris-Aligned Benchmarks (PAB) and, in particular, in the Euronext SBT-1.5° CAC" (p.98).

Back-filled from the climate risk mapping in section 3.3.1.4 of the FY2025 report (pages 97-100). This disclosure requirement did not exist under the 2023 ESRS the report was prepared against.

Reference: pages 97-100

Risk identification. Risks are separated by origin, "A) Direct operations... B) Supply chain" (p.97), and then by type. Acute and chronic physical risks are tabled on p.99, and transition risks on pp.99-100 under four headings: existing regulations, emerging regulations ("Carbon taxes, air travel taxes and mandatory energy audits"), technological risks and market risks. Each carries an impact, a mitigation strategy, a horizon and a criticality rating, with technological risk the only one rated high.

Scenarios. "The analysis is based on the IPCC's SSP2-4.5 and SSP5-8.5 scenarios. The SSP2-4.5 scenario is the intermediate scenario, forecasting global warming of around 2°C medium-term and 2.7°C long-term over pre-industrial levels. The SSP5-8.5 scenario is the most pessimistic, forecasting global warming of 2.4°C medium-term and 4.4°C long-term over pre-industrial levels" (p.98). A high-emission scenario is therefore used and temperature projections are given. No 1.5°C-aligned scenario is named for transition risk.

Scope. The exercise is "based on the location of commercial operations", with country results for India, the Philippines, Colombia, Brazil, Mexico and Egypt across seven hazards (p.99). The date of the analysis is not stated.

Back-filled from the climate risk vulnerability analysis in section 3.3.1.4 of the FY2025 report (page 98) and the mitigation columns of the physical risk table (page 99). This disclosure requirement did not exist under the 2023 ESRS the report was prepared against.

Reference: pages 98-99

The conclusion is stated directly and quantified by asset exposure: "According to the climate risk vulnerability analysis thus developed, the Group's business model is resilient to climate change; 15% of its office space is located in a medium-to-high vulnerability zone, 14% in a low-to-medium vulnerability zone and the remaining 71% in a low vulnerability zone" (p.98). The concentration is named: "Medium to high vulnerability locations include India and the Philippines, where TP has about 30% of its workforce, and to a lesser extent Colombia, Brazil, Mexico and Egypt."

Adaptive capacity appears in the mitigation column of the physical risk table (p.99): business continuity plans at the most exposed subsidiaries, mitigation "by the Group's geographic diversification", and "Contractual business continuity plans include the rollout of emergency solutions and alternative means of production".

Two elements have nothing to draw on: significant areas of uncertainty are not identified, and there is no analysis of financial flexibility or of the capacity to redeploy, repurpose or decommission assets. The report does not say when the analysis was last refreshed.

Reference: pages 97, 99, 106

"TP's commitment is underpinned by an environmental policy targeting two main objectives: reducing environmental impact and raising employee and stakeholder awareness" (p.97).

Scope and content are stated: "The policy covers all operations and transactions, including employees, subcontractors, suppliers and third parties. It provides for climate change mitigation and adaptation measures. It also includes measures to improve energy efficiency and promote sustainable solutions, as well as a commitment to deploy renewable energies across the Group, responsible procurement and the circular economy" (p.97). Both mitigation and adaptation are therefore covered by a single instrument, and the policy also carries the E5 circular economy content.

External reference points: "This policy is in line with the principles of the United Nations Global Compact, the Sustainable Development Goals (SDGs) and the Science-Based Targets initiative (SBTi)" (p.97), and "The Group's environmental policy and environmental performance management system are based on the principles of the ISO 14001 standard" (p.99). 48% of operations hold ISO 14001 certification (p.109).

Not disclosed: the policy's approval date or review cycle, and whether energy efficiency, renewable deployment and climate adaptation are covered by separate documented standards beyond the Global Premises Standard referred to on p.99.

Reference: pages 100-105

Actions are grouped by lever and, unusually, quantified as a bridge from the 2019 baseline (p.101). The decarbonization lever chart runs from 723,408 tCO2e in 2019 through business growth of +142,781 and reductions from renewable energy procurement (120,135), energy efficiency (32,179), site optimization (37,230), work-at-home (120,402), commuting changes (12,329), supplier engagement (67,270), business travel (5,373) and behaviour (19,272) to the 2030 target of 451,995 tCO2e.

Energy efficiency (p.104): LEED-compliant high-performance buildings, subsidiary energy performance reviews, an energy management system, low-energy light bulbs, motion detectors, light sensors and timers, air-conditioning optimisation, and STAR- and EPEAT-certified equipment. A best practices guide splits measures into "Must-Have" initiatives for all facilities and "Invest with Rapid Payback" initiatives assessed on payback period.

Commuting and travel (pp.101, 104-105): TP Cloud Campus, public transport subsidies, shuttle contracts used by "around 54,000 employees", bicycle parking, carpooling; and for travel, a strengthened policy, video-conferencing and "prefers rail to air for journeys under four hours". A simulator built with the CESG quantifies the Cloud Campus effect: "a work-from-home advisor generates 55% fewer GHG emissions per year than an advisor working on site" (p.101).

Reference: pages 96, 98, 103

The targets are absolute, science-based and validated. "TP is committed to reducing Scope 1 and Scope 2 emissions by 56.7% by 2030, compared to the 2019 baseline, and reducing Scope 3 emissions by 27.5% over the same period, both in absolute terms" (p.98). The Scope 1 and 2 target is "compatible with the scenario of limiting global warming to 1.5°C" and was approved by the SBTi in 2024, replacing a 2021 set of targets aligned to below 2°C (p.97). Base year 2019.

An energy target sits alongside: renewable energy at "50% by 2026 and 80% by 2030" as a share of total electricity consumption (p.96).

Progress against target is published rather than left to the reader. The emissions table on p.103 carries a "% achievement to date versus targets" column: 87% for the Scope 1 and 2 target and 37% for the Scope 3 target. "Thanks to the measures implemented, TP is ahead of its Scope 1 and 2 decarbonization targets. Scope 3 emissions are also declining, in line with targets, although the ongoing trend of returning to the office and the Group's business growth may slow these effects."

Gaps worth naming: there is no 2050 net-zero target. Page 98 records the 2030 targets as validated for all scopes, "with the willingness to work on a 2050 target", and no milestone is set between 2025 and 2030.

Reference: pages 103-104

The full energy table is given for 2023, 2024 and 2025 (p.104). For 2025, in MWh: coal and coal products 0; crude oil and petroleum products 22,393; natural gas 8,761; other fossil sources 0; purchased electricity, heat, steam and cooling from fossil sources 205,195. Total fossil energy consumption 236,348 MWh, 53.44% of the total (down from 56.79% in 2024). Nuclear 365 MWh, 0.08%. Renewable fuel 0; purchased renewable electricity, heat, steam and cooling 203,772; self-generated non-fuel renewable 1,779; total renewable energy consumption 205,551 MWh. Total energy consumption 442,264 MWh, down from 458,910 in 2024.

Electricity consumption is reported separately at 411,110 MWh, down 3% on 422,878 MWh in 2024 (p.104), split 50.0% renewable, 49.9% non-renewable and 0.1% carbon-free. "Renewable sources of energy primarily consist of solar, wind and hydro. 37% of the renewable energy used comes from renewable energy certificates (EAC contractual instruments), which represent 19% of the Group's total electricity consumption."

Two renewable percentages appear and they measure different things, which is worth flagging to a reader comparing them. The table row reported as 46.48% is total renewable energy over total energy consumption (205,551 / 442,264), while the 50.0% in the narrative is renewable energy over the 411,110 MWh of electricity consumed. Both are consistent with the printed figures.

Reference: pages 102-103

The 2025 footprint, market-based, is 551,700 tCO2e, against 602,799 in 2024 and 723,408 in the 2019 baseline: down 7% year on year and 24% against 2019 (p.103). Location-based, the total is 639,477 tCO2e.

Scope 1: 21,997 tCO2e (-5%), comprising stationary fuel combustion 4,765, company vehicles 2,687 and refrigerant leakage 14,546. "% of Scope 1 emissions resulting from regulated emissions trading" is reported as 0.

Scope 2: 102,601 tCO2e market-based (-1%) and 190,378 location-based (+1%). Scope 1 and 2 combined are 124,599 market-based, 50% below the 2019 baseline, and 212,376 location-based.

Scope 3: 427,101 tCO2e (-10%): purchased products and services 182,625 (-35%), capital goods 90,955 (+77%), commuting 134,860 (+6%), business flights 17,037 (+13%) and other business travel 1,625. Three further categories sit outside the reduction targets and are reported separately: waste 1,260, upstream transportation of goods 4,867 and energy-related emissions not in categories 1 and 2 at 7,523.

Boundary decisions are tabled on p.102. Categories 1, 2, 6 and 7 are "Relevant, included" and represent 97% of Scope 3; categories 3, 4 and 5 are "Relevant, non-material" at under 3%; categories 8 to 15 are "Not applicable" because "TP is a service company that does not sell any products. Moreover, the Group has no upstream or downstream leased assets, downstream franchises or investments external to the Group."

Reference: pages 98, 100

A nil return, stated in one sentence: "TP does not use carbon credits to offset its emissions" (p.100). The Appendix B cross-reference table confirms the placement, listing "ESRS E1-7 GHG removals and carbon credits, paragraph 56" against section 3.3.1.5 (p.139).

Consistent with that, the targets are described as absolute reductions rather than net figures: Scope 1 and 2 down 56.7% and Scope 3 down 27.5% by 2030 against the 2019 baseline, "both in absolute terms" (p.98), and the decarbonization lever bridge on p.101 contains no removal or offset lever.

The Group does describe restoration activity, but does not claim removals from it. Under "HOW DO WE CONTRIBUTE TO RESTORATION?" the report lists "Partnerships with organizations specializing in reforestation", "Biodiversity conservation campaigns" and "World Cleanup Day" (p.98), and section 3.1.2 records that "TP supports reforestation programs in its key countries" (p.77). No tonnage of removals is attributed to those programmes, no carbon credit is cancelled or retired, and no removal project in the own operations or the upstream or downstream value chain is quantified.

Reference: page 100

A nil return with a stated alternative: "The Group does not apply internal carbon pricing, but sets targets for each subsidiary based on the subsidiary's carbon emissions in absolute terms and per employee" (p.100).

No shadow price, internal fee or carbon price per tonne is disclosed for any scheme, and no carbon price is applied in investment appraisal or in the CapEx and OpEx allocation described in the same paragraph. The subsidiary-level mechanism the company puts in its place is a target-setting and allocation device, not a price: emissions budgets are pushed down to each subsidiary in absolute terms and per employee, and progress against them feeds the Group trajectory.

Related incentive mechanisms exist but are also not carbon prices. A "CSR" environment criterion weighted at 12.5% of the 2025 performance share plan is based on the rate of Scope 1 and Scope 2 reduction against the 2030 targets (p.85), and the Group's credit facilities and its 2022 bond carry criteria on GHG emission reduction and renewable energy adoption (p.85). These attach money to outcomes rather than setting a price per tonne of carbon.

Reference: pages 98-100, 107-108, 139

The company treats E1-9 as covered. The Appendix B cross-reference table (p.139) maps four E1-9 datapoints to report sections: exposure to physical risks (paragraph 66) and the disaggregation by acute and chronic risk and location of significant assets (paragraph 66(a) and (c)) to section 3.3.1.4; the carrying value of real estate assets by energy-efficiency class (paragraph 67(c)) to section 3.3.3; and exposure to climate-related opportunities (paragraph 69) to section 3.3.1.4.

What those sections contain is asset exposure and qualitative effect, not money. Physical risk exposure is quantified by share of estate: "15% of its office space is located in a medium-to-high vulnerability zone, 14% in a low-to-medium vulnerability zone and the remaining 71% in a low vulnerability zone" (p.98), with a country hazard table on p.99. Effects are described in words: "Reduction in TP revenues due to temporary shutdown of operations" and "rehabilitation costs after a climate event and business disruptions" (pp.91, 99). On real estate, the Taxonomy analysis reports EUR 61.5 million of aligned CapEx, 6.3% of the EUR 984.6 million total (pp.107-108).

No monetary amount is disclosed for anticipated financial effects from physical risks, transition risks or opportunities, and no carrying value of assets at material physical risk is given.

Reference: pages 105-106

The circular economy content sits inside the single environmental policy, which "is under the supervision of the Board of Directors" (p.106). "The Group applies the principles of circularity for waste management, focusing on reduction at source, sorting, recycling and responsible disposal. TP measures the amount of waste generated, limits packaging and single-use plastics, and encourages the refurbishment, recycling or donation of electronic equipment at the end of its life" (p.106).

A second instrument is named in the material topic table: the KPI table for e-waste management lists "Environmental policy; IT equipment management policy" as the policies in place (p.105).

Green IT is the operative commitment: "TP is committed to reducing the environmental footprint of electronic equipment through Green IT initiatives, including waste control, participation in take-back programs and obtaining regulatory-compliant treatment certificates. Key performance indicators have been expanded to enhance visibility and continuous improvement of the equipment life cycle" (p.106).

The policy addresses resource outflows and waste. It does not set out an approach to resource inflows or to sustainable sourcing of materials, consistent with the finding that "responsible sourcing of materials" is not material (p.91).

Reference: page 106

"TP has introduced various standardized processes for the collection, sorting and disposal of waste generated by its activities. The overall strategy adopted for waste management at each TP facility includes the following three elements: waste reduction at source; waste sorting at source; reuse, recycling and disposal" (p.106).

Three action lines are described.

Supplier take-back. "TP also engages its suppliers through various waste reduction initiatives such as take-back programs, in which suppliers are required to take back IT hardware such as computers, headsets, monitors and workstations and take over responsibility for processing these devices."

Life extension. "TP strives to extend the life cycle of electronic equipment through internal resource reallocation and the maintenance and refurbishment of existing equipment", and the material IRO table records the countervailing financial logic: "Reducing costs by extending equipment lifespan and investing in durable equipment" (p.91).

Employee engagement. "Employees receive regular communications in the form of e-mail, posters and campaigns."

No monetary resources are attached to any of these actions, and no implementation timeline is given.

Reference: page 105

One target is stated, and it is qualitative. The material topic table for ESRS E5 sets out, for the KPI "Electronic waste donated or recycled (in tonnes)", the target "Increase the proportion of recycled IT assets" (p.105), with the associated policies given as the environmental policy and the IT equipment management policy.

The measured base against which that direction of travel runs is disclosed: e-waste donated or recycled fell from 771 tonnes in 2024 to 549 tonnes in 2025, with no figure available for 2023 (p.105). The underlying treatment split on p.106 shows the movement more precisely: of 1,284 tonnes of e-waste generated in 2025, 41% was recycled and 2% donated, against 62% and 8% of 1,097 tonnes in 2024.

The company qualifies the comparison itself: "At this stage, year-on-year comparability remains limited and will require further maturation of the reporting process" (p.106).

Judged against what ESRS E5-3 asks for, this is a directional commitment rather than a measurable target: there is no quantified value, no target year and no base year, no separate target for waste reduction, recycled content or resource inflows, and no statement of whether stakeholders were involved in setting it. The Group's only quantified environmental targets are the climate and renewable energy targets reported under E1-4.

Reference: page 106

Outflows are reported as electronic waste by destination, which is where the Group's material IRO sits. "The table below presents TP's best estimates of the tonnage of electronic waste generated, broken down by sorting method" (p.106).

For 2025, 1,284 tonnes of e-waste generated (1,097 tonnes in 2024), split: reused or repaired 97 tonnes (8%, against 211 tonnes and 19% in 2024); recycled 530 tonnes (41%, against 683 tonnes and 62%); donated 19 tonnes (2%, against 88 tonnes and 8%); incinerated 0 tonnes (0%); and disposed of in accordance with applicable standards 638 tonnes (50%, against 114 tonnes and 10% in 2024). The movement between recycling and disposal between the two years is large and is not explained beyond the company's own caution that "year-on-year comparability remains limited and will require further maturation of the reporting process".

The IRO-2 Appendix B table maps two datapoints here: "ESRS E5-5 Non-recycled waste, paragraph 37 (d)" and "ESRS E5-5 Hazardous waste and radioactive waste, paragraph 39", both to section 3.3.2.3 (p.140).

Not disclosed: no split of hazardous against non-hazardous tonnage, no products-and-materials outflow figures, and no expected durability, repairability or recyclability characteristics of the services TP sells.

Reference: pages 105-106

Waste is disclosed as a single stream, electronic waste, because that is where the material IRO sits: "The management of electronic waste (e-waste) is a material topic for TP. The regular upgrading of IT equipment, which is necessary in order to remain at the cutting edge of innovation, generates specific waste. Although the volume per employee is limited, TP acknowledges that unsuitable practices could have consequences for public health and the TP's image" (p.105).

Total waste generated in 2025: 1,284 tonnes of e-waste, up from 1,097 tonnes in 2024. By treatment route (p.106): reused or repaired 97 tonnes (8%); recycled 530 tonnes (41%); donated 19 tonnes (2%); incinerated 0 tonnes (0%); disposed of in accordance with applicable standards 638 tonnes (50%). Waste diverted from disposal - reuse, repair, recycling and donation combined - therefore accounts for 646 tonnes, 50% of the total, against 89% in 2024.

The headline KPI in the material topic table reports the diverted portion differently again: "Electronic waste donated or recycled (in tonnes)" of 549 tonnes in 2025 against 771 tonnes in 2024 (p.105).

Not disclosed: the tonnage split between hazardous and non-hazardous waste, radioactive waste (none is claimed), waste by composition, or the amounts sent to landfill against other disposal routes within the 50% "disposed of in accordance with applicable standards" line.

Reference: pages 110-111, 114, 118, 122, 130

Policies are described per material sub-topic rather than as a single own-workforce policy. "Its policies, which are regularly updated, cover the IROs identified by the double materiality analysis: working conditions, equal treatment, social dialog, health and safety, training, career development, privacy and data security" (p.111).

Human rights policy (p.114): "aligned with the main international standards, including the United Nations Global Compact, the Universal Declaration of Human Rights, ILO conventions and the OECD Guidelines... The policy covers key topics such as discrimination, harassment, workplace safety and the prevention of human trafficking, forced labor and child labor. It also covers working hours, minimum wages, freedom of association, collective bargaining, privacy, freedom of expression, wellbeing at work and the protection of mental health, as well as the ethical use of artificial intelligence." It "applies to all Group business activities and stakeholders" and "provides for whistleblowing mechanisms, including a Global Ethics Hotline". The Appendix B table maps the paragraph 20, 21 and 22 datapoints on human rights commitments, ILO conventions 1 to 8 and trafficking to section 3.4.1.5 (p.140).

Data security policy (p.130), covering employees among others, with Global Information and Security Policies and global privacy and compliance standards listed item by item.

Reference: pages 111, 115-116, 121-122

"TP has established structured processes to manage its impacts, risks and opportunities, promoting regular interaction with employees and their representatives. Social dialog is conducted with trade unions and representative bodies at all levels of the Group and is adapted to local specificities and legislation in force. Tailored engagement strategies are deployed at each facility, involving local stakeholders in decision-making" (p.111). Employees are represented by two directors on the Board.

Scale and outcome are given. "In 2025, 262,665 employees submitted confidential replies to the independent employee trust surveys conducted by the Great Place to Work Institute to measure their trust in their employer (63% response rate)"; the average trust index was 77%, certification was obtained in 69 countries covering 90% of the workforce, and TP "ranked seventh among the World's 25 Best Workplaces across all industries" (p.116).

The closed loop is described: "the results are analyzed daily by the HR teams then passed through weekly and monthly reviews in order to adjust action plans and ensure their effectiveness" (p.116), with country HR managers accessing local results and a global wellbeing team coordinating action plans against defined indicators (participation rate, Net Happiness Score target 55%, average score target 4.5, Top 2 Box target 85%, Bottom 3 Box maximum 15%).

Reference: pages 114, 121, 134

The main channel is Group-wide and open beyond employees: "The Group provides a secure and confidential hotline that allows any employee or stakeholder, internal or external, to report any suspected violation of ethical standards or the code of conduct: corruption, antitrust practices, human rights violations, discrimination, environmental damage, health and safety breaches, fraud, or any inappropriate professional conduct. This system complies with local legislation and, where applicable, is backed up by local whistleblowing mechanisms" (p.134). The Appendix B table maps "ESRS S1-3 Grievance/complaints handling mechanisms, paragraph 32 (c)" to section 3.5.2 (p.140).

Independence and handling are described: "Alerts are handled by a dedicated team attached to the Legal and Compliance Department, thereby ensuring the independence of the process. The employees responsible for receiving and handling alerts are formally appointed and receive specific training... Investigations follow the procedures defined in the Global Ethics Hotline Policy. Regular reports are presented to the Audit, Risk and Compliance Committee, while major cases are forwarded to senior management. Corrective measures may include disciplinary sanctions, process improvements and targeted audits" (p.134).

Protection against retaliation is stated: "No form of retaliation is tolerated against persons who submit alerts in good faith" (p.134).

Reference: pages 113-126

Actions are organised by the five material sub-topics.

Working conditions. Living-wage benchmarking with Wage Indicator covering 99% of the workforce in 107 subsidiaries (p.117); a global broker selected at the end of 2025 to take over supplementary health, death, accident, disability and incapacity cover across all countries, with transition due by end-2026 (p.117); free shuttle services for around 54,000 employees where public transport is inadequate or for night work (p.118); TP Cloud Campus teleworking, used by around 30% of employees (p.118).

Human rights. Manager training completed by "more than 6,800 managers... giving a completion rate of over 80%"; 54 subsidiaries assessed since inception covering 92% of the workforce and all countries most at risk on the Human Rights Index Score; and a substantially enlarged social audit function: "In 2025, the Group significantly strengthened its social audit function, extending coverage to 15 countries and organizing 315 confidential roundtables involving around 3,000 employees at 54 facilities. The introduction of the Social Risk Exposure (SRE) assessment framework was a major step forward" (p.115).

Equal treatment. TP Women mentoring, Best Workplaces for Women certification in nine countries, and Impact Sourcing programmes for refugees, young people without diplomas and persons with disabilities; "Over 6,200 persons with disabilities were recruited in 2025" (pp.119-120).

Reference: pages 110, 123

Targets are set out in the material topic table at the head of the S1 section, with the 2023-2025 series against each (p.110):

• Employees working in an environment recognised as a Great Place to Work: 99%, 97%, 90% against a target of "Over 90% each year".
• Employee satisfaction rate in the Great Place to Work surveys: 79%, 78%, 77% against "Over 70% each year".
• Women in the workforce: 53.7%, 53.0%, 52.6%, and women in management positions 51.9%, 48.2%, 48.8%, both against "Maintain gender balance > 45%".
• Employees covered by an ISO 45001 health management system: 60%, 80%, 99.5% against "100% in 2025".
• Employees trained in the Health & Safety Policy: 94%, 90%, 88.4% against "Over 90% each year", the shortfall attributed to a change of training year to April-March.
• Number of training hours per employee: 138, 135, 134, against "Continuous training opportunities for all employees".
• Internal promotion rate: 61%, 72%, 68% against "Over 60% each year".

The KPI dashboard on p.87 rates each of these as ahead of target, on target or "More action needed".

One deliberate absence is explained. On health and safety metrics, "The Group does not wish to establish quantitative objectives in order to encourage subsidiaries to be exhaustive in their reporting" (p.123), so there is no accident frequency or lost-day target.

Reference: pages 110, 112-113

Headcount at 31 December 2025 was 485,606, against 489,488 a year earlier, with a further "approximately 16,000 non-employees" reported separately (pp.112-113).

By gender: men 227,128 (46.8%), women 251,867 (51.9%), other 12, not reported 6,599. By age: under 30 278,857; 30 to 50 182,789; over 50 23,960. By region: Americas 290,701; Europe, MEA and Asia Pacific 176,813; with segment splits for Core Services 468,230, Specialized Services 17,266 and holding companies 110 (p.112). The top ten countries are tabled, led by India 94,656, the Philippines 55,449, Colombia 44,639, the United States 30,434 and Egypt 24,731.

By contract type (p.112): permanent 391,622, fixed-term 77,053, temporary 16,931; full-time 424,427 and part-time 61,179. The 2025 contract table is also broken out by region on p.113.

Full-time equivalents: 446,716 against 446,052 in 2024, with payroll expenses of EUR 6,950 million (p.112).

Movements are given rather than only closing balances (p.113): hiring 361,027, resignations -252,340, lay-offs -75,318, other departures -27,814, a net change of -3,882. "In 2025, in accordance with CSRD requirements, the turnover rate stood at 73%, compared with 80% in 2024. It includes all voluntary and involuntary departures." The Group also reports its own attrition measure: advisers, over 80% of the workforce, averaged 5.6% per month, 67% for the year, and the Group average was 5% per month or 60% for the year.

Reference: page 122

Coverage is reported in bands rather than as a single percentage, using the format the standard allows for regions above 10% of the workforce (p.122):

• EEA employees: collective bargaining coverage 80-100%, and workplace representation 80-100%. "No single country in the EEA accounts for more than 10% of the Group workforce."
• Non-EEA regions above 10% of employees: North America, India and the Philippines at 0-19%; EMEA and Asia-Pacific excluding the EEA, and Latin America, at 20-39%.

Above the local arrangements sits a global agreement: "Since December 2022, TP has had a framework agreement with the UNI Global Union (UNI) international trade union federation, thereby strengthening its commitments to upholding employees' rights to form trade unions. TP also recognizes UNI as a stakeholder under the French duty of vigilance law. This agreement covers 100% of the Group's employees" (p.122). Separately, "local unions are recognized in 24 countries, covering 39% of the Group's employees".

Social dialogue structures are described: staff representatives, works councils, health and safety committees and grievance committees, and "In countries where these fundamental freedoms are not guaranteed, TP ensures that channels for social dialog exist" (p.121). A European Works Council of 19 standing members and one observer represents EEA employees, and its 2025 agenda is listed.

Reference: pages 112, 119

Gender distribution is reported at four levels of the organisation, with 2024 comparatives and the Group objective against each (p.119):

• Board of Directors: 6 women and 7 men at 31 December 2025, women 46.2%, against 50% in 2024; objective over 40%.
• Executive Committee: 3 women and 5 men, women 38%, against 30% in 2024; objective over 40%.
• Management (defined as "all employees except agents and supervisors"): 25,848 women and 27,114 men, women 48.8%, against 48.2% in 2024; objective over 45%.
• Total headcount: 251,867 women and 227,128 men, women 52.6%, against 53.0%; objective over 45%.

The Executive Committee figure is the one sitting below its stated objective.

Age distribution for the whole workforce is given in the S1-6 table on p.112: under 30, 278,857; 30 to 50, 182,789; over 50, 23,960 - so 57% of employees are under 30. The report does not break that distribution down for top management.

Board gender diversity is also the subject of an Appendix B datapoint, mapped to section 4.1.3 (p.139), and the Board composition table on p.83 identifies each director's gender and independence individually.

The auditor gave S1-9 particular attention alongside S1-6 (p.144).

Reference: page 117

The legal-minimum test is answered without qualification: "TP ensures that all of its employees receive remuneration in excess of local legal requirements. As in 2024, in all countries where TP operates, 100% of employees receive a salary above the statutory minimum, in line with social compliance standards" (p.117).

The Group then goes beyond the minimum-wage test to a living-wage benchmark, and names the provider and the coverage. "Aware that the minimum wage does not always reflect the real cost of living, TP has strengthened its commitment to offering responsible remuneration by partnering with Wage Indicator, a benchmark organization in living wage analysis. This partnership makes it possible to assess the adequacy of the salaries paid by TP by comparison with the living wage in each geographical region, by means of an annual analysis covering 99% of the Group workforce in 107 subsidiaries."

What is not published is the result of that annual analysis: no percentage of employees paid at or above the living wage in any country or region is given, so the reader is told the benchmark exists and how wide it reaches, but not what it found.

Reference: page 120

A figure is given, as a range, with the reason for the imprecision set out first: "The percentage of employees with disabilities is reported in the Group's staff data consolidation tools, in compliance with privacy and data protection regulations. In some subsidiaries, local law prohibits the collection of this information. Similarly, employees are not required to declare this personal information to their employer and do so on a voluntary basis. Therefore the data collected is partial" (p.120).

To compensate, "the Group also relies on the results of the confidential annual 'Census' survey, to which employees respond on a voluntary basis. On this basis, the Group estimates that persons with disabilities accounted for between 2% and 4% of the workforce in 2025, stable versus 2024."

Recruitment volume is disclosed: "Over 6,200 persons with disabilities were recruited in 2025", supported by adapted hiring processes, facilities accessible to persons with reduced mobility, and partnerships with specialised associations. Persons with disabilities are one of the five priority groups in the Impact Sourcing programme, with named country initiatives in India and the Philippines, including Project Echo for the deaf and hard of hearing (pp.120-121).

What is not given is a single Group percentage on a consistent basis, or a breakdown by gender or region.

Reference: pages 110, 125-126

Training volume is reported in both absolute and per-head terms: "In 2025, 59.6 million training hours were provided, representing 134 hours per full-time equivalent (135 in 2024)" (p.125). The material topic table gives the three-year series - 138, 135 and 134 hours - against a target of "Continuous training opportunities for all employees" (p.110).

Coverage is stated to be universal: "All employees, including part-timers, temps and subcontractors, benefit from the training" (p.125), and "All new employees attend an induction seminar on their first day, with a strong focus on Group culture and values."

Volume by programme is given for the priority areas (p.126): "In 2025, over 313,000 people were trained in EI and around 65,000 in AI (total: 120,000+). A new AI cycle started in January 2026"; the language academy had "28,047 learners in 2025 (vs. 23,865 in 2024)"; around 9,700 employees took part in the Jump! programme; and satisfaction was "95% for online training and 96% for in-person training", with mental health training at a 92% satisfaction rate.

The gap is named by the company: "Gender breakdown is not yet available, but the Group plans to improve monitoring on this point" (p.125). No percentage of employees who actually received a performance review is given either, only the design of the process.

Reference: pages 110, 122-123

Coverage of the management system is reported at two levels. 99.5% of the Group's entities are ISO 45001 certified, up from 80% in 2024 (p.123), against a stated target of 100% in 2025, and the material topic table reports "Percentage of employees covered by an ISO 45001 health management system" at 60%, 80% and 99.5% across 2023 to 2025 (p.110). Health and safety policy training reached 88.44% of employees between 1 April and 31 December 2025, with the remainder due to complete by 30 March 2026; the year-on-year fall is attributed to the training year moving from calendar to April-March.

The incident metrics are tabled (p.123), 2024 against 2025:

• Number of workplace accidents recorded: 1,570 to 1,064
• Proportion of workplace accidents recorded: 0.32 to 1.06
• Number of occupational illness cases recorded: 49 to 76
• Number of deaths due to workplace accidents and occupational illnesses: 7 to 3
• Number of days lost due to workplace accidents, death, other accidents or ill health: 18,178 to 11,535
• Workplace accidents and occupational illnesses among other employees working at TP facilities: 0; deaths among them: 0

The rate line moves in the opposite direction to the count, from 0.32 to 1.06, and the report does not explain the change of basis. The Appendix B table maps the fatality and accident-rate datapoints (paragraph 88(b) and (c)) and days lost (paragraph 88(e)) to section 3.4.1.8 (p.140).

Reference: page 118

Both limbs of the metric are reported, for 2024 and 2025, split by gender (p.118):

• Percentage of employees authorized to take maternity/paternity leave: women 100%, men 99%, total 99% in 2025, against 100%, 94% and 97% in 2024.
• Percentage of employees who took maternity/paternity leave: women 0.6%, men 0.2%, total 0.4% in 2025, against 3%, 1% and 2% in 2024.

The collection basis is stated: "The data is collected on local HR management systems and reported to the Group through an annual internal survey sent to all the subsidiaries", and "Family leave arrangements vary according to local regulations in each country of operation."

The scope limitation is disclosed twice, in the same words in the S1 section and in the methodology note: "TP does not publish information relating to parental or carer's leave. The necessary data is not available homogeneously within HR systems, and national practices differ significantly from one country to another, not allowing for reliable consolidation at Group level at this stage" (pp.118, 138). So the entitlement and take-up figures cover maternity and paternity leave only.

Reference: page 119

Both required ratios are given, with a comparative and an unusually candid note on methodology.

Gender pay gap: "In 2025, the average gender pay gap at TP was 2.7% in favor of men, representing a slight decrease compared with 2024" (p.119).

Total compensation ratio: "The ratio between the highest pay and employee salaries was 506. Despite the distribution of the workforce and the Group's extensive international footprint, TP applied no weighting according to the workforce of each country or any adjustment related to parity in cost of living, which significantly increases the published gap. In addition, the denominator includes part-time employees. It amounted to 655 in 2024, due to the higher valuation of free share awards."

That is a fall of 149 points year on year, and the company attributes the 2024 level to share award valuation rather than to a change in pay policy. The disclosure that no cost-of-living or workforce weighting was applied is helpful for comparability: it tells a reader that the figure is a raw ratio across around 100 countries, most of the workforce being in India, the Philippines, Colombia and Egypt.

The Appendix B cross-reference table lists both datapoints - "ESRS S1-16 Unadjusted gender pay gap, paragraph 97 (a)" and "ESRS S1-16 Excessive CEO pay ratio, paragraph 97 (b)" - against section 3.4.1.6 (p.140).

Reference: page 134

Incidents are reported through the Global Ethics Hotline table, which covers own workforce and value chain alike (p.134). Volume first: "In 2025, 3,340 reports were received, compared with 1,643 in 2024. 2,450 (73%) fell within the scope of the ethics hotline, while the other alerts mainly concerned administrative or organizational matters (e.g. payroll, schedule management) and were referred to the relevant departments."

By type, with alerts received, alerts substantiated and fines, sanctions and compensation related to corruption or serious human rights violations in euros:

• Incidents of discrimination, including harassment: 1,218 alerts, 334 substantiated, EUR 0
• Serious human rights incidents: 0 alerts, 0 substantiated, EUR 0
• Corruption incidents: 13 alerts, 3 substantiated, EUR 0
• Data privacy incidents: 35 alerts, 17 substantiated, fines "Not applicable"
• Other types of incidents: 1,184 alerts, 414 substantiated, fines "Not applicable"

The doubling of reports year on year is not explained in the text, and the report does not say whether it reflects more incidents or more use of a better-publicised channel.

The Appendix B table maps "ESRS S1-17 Incidents of discrimination, paragraph 103 (a)" and "ESRS S1-17 Non-respect of UNGPs on Business and Human Rights principles and OECD guidelines, paragraph 104 (a)" to section 3.5.2 (p.140).

Reference: pages 127-128

The governing instrument is the Supplier Code of Conduct. "TP requires its suppliers to sign a code of conduct that clearly defines the requirements with regard to human rights, working conditions, health and safety, the environment, data protection, ethics and integrity. Based on five values (integrity, respect, innovation, professionalism and commitment), the Code requires compliance with international standards, in particular those of the ILO on working time and conditions" (p.128).

Commitments are itemised on p.128: human rights; non-discrimination; health and safety standards; freedom of association and collective bargaining; data protection; "Responsible use of resources and prevention of pollution"; and "Prohibition of child labor and forced labor". The Code "is accessible to all stakeholders on the Group's website and intranet".

Accountability sits with the Global Procurement Department, which reviews the Supplier Code of Conduct, and with "A Third-Party Risk Committee representing the procurement, sustainability, personal data, information security, and legal and compliance functions", which "oversees the implementation of due diligence measures and the monitoring of supplier risks, reporting regularly to the Global Compliance and Security Council and senior management" (p.127). "Following an external evaluation, the procurement policy was strengthened in order to formalize due diligence procedures."

Reference: pages 86, 127-128

Operationally, engagement takes three forms (pp.127-128). Structured supplier monitoring: "A structured system is used to monitor and evaluate working conditions, including regular interactions and specific supplier assessments, in order to promote human rights and identify areas for improvement." Documentary assessment: high-risk suppliers complete an assessment on the Integrity Next platform covering environmental, social and governance requirements. Direct on-site work with the suppliers whose workers are physically present: "Extensive audits are regularly carried out with strategic service providers working on site (maintenance, security, catering, transport of employees, etc.), with a particular focus on working conditions and compliance with health and safety standards. In addition, specific training is provided to external operators to ensure they fully understand the applicable minimum requirements."

Risk targeting is described: "The risks related to working conditions and human rights in the value chain are analyzed as the combination of sectoral risk and gross country risk as established by the Human Rights Index Score" (p.127), with four purchasing categories mapped to their characteristic impacts, including "Living wage, working hours, leave" for temporary employment agencies.

Not disclosed: whether value chain workers or their representatives are consulted directly, and how the effectiveness of engagement is assessed.

Reference: page 128

The channel is the same Global Ethics Hotline used by employees, opened explicitly to the value chain. "TP has set up an ethics whistleblowing system accessible to all employees, including value chain stakeholders, in order to guarantee the protection of rights and promote a responsible working environment. The Global Ethics Hotline available on www.tp.integrityline.com allows them to report any inappropriate behavior or human rights violations confidentially and without fear of retaliation" (p.128).

Handling and protection are described: "Alerts are processed rigorously and quickly, and are subject to thorough investigations. TP commits to taking the necessary corrective measures and to protecting the identity of the persons concerned. Internal policies, including the Code of Conduct and the Global Ethics Hotline Policy, ensure the protection of whistleblowers acting in good faith. A detailed report on alerts submitted and action taken is published in the Universal Registration Document."

The outcome for the year is stated as a nil return: "In 2025, there were no reported cases in the value chain" (p.128). Alerts overall totalled 3,340, of which 2,450 fell within the hotline's scope, and serious human rights incidents were reported as zero across the whole system (p.134).

Reference: pages 127-128

The action is a graduated due diligence funnel, described stage by stage and, at the final stage, quantified (p.128).

• Preliminary risk assessment: "Each new supplier undergoes initial analysis via a criticality questionnaire completed by the purchasing teams. This assessment takes into account the business vertical, the country of establishment, the degree of dependency and the amount of expenditure incurred."
• Classification by risk: "Suppliers with a medium or high gross risk are subject to an assessment covering the following themes: anti-corruption, respect for human rights, health and safety, the environment, and privacy, as well as verification of sanctions lists."
• Non-compliance management: documentary or on-site audits and corrective action plans, with contract termination "as a last resort".
• In-depth documentary audits for high-risk suppliers: "In 2025, 596 suppliers were identified as subject to high gross risk; all received a request for assessment. 20% of them had completed their assessment by December 31, 2025. The aim is to assess 75% of at-risk suppliers by 2026."
• Audits and training for strategic suppliers working on site in maintenance, security, catering and employee transport, "with a particular focus on working conditions and compliance with health and safety standards".

Reference: page 127

Two targets are set out in the material topic table at the head of the ESRS S2 section, each paired with the KPI that measures it (p.127):

• "Deployment of the Group due diligence procedure", reported at 100% for 2023, 2024 and 2025, against the target "Roll out the procedure across all Group entities".
• "Proportion of at-risk suppliers having completed an in-depth assessment of their CSR practices", reported at 20% for 2025 with no comparable figure for earlier years, against the target "Assess the CSR practices of 75% of suppliers identified as high risk by 2026".

The second is a dated, quantified target with a measured starting point: 596 suppliers were identified as high gross risk in 2025, all were asked to complete an assessment, and 20% had done so by 31 December 2025 (p.128). Reaching 75% within a year is a substantial step-up and the report does not describe how it will be achieved.

Comparability is limited by a change of method, which the company footnotes: for the at-risk supplier KPI, "As the methodological approach has evolved, the previous data cannot be restated" (p.127).

Reference: pages 129-130

The material S4 topic is privacy and data security for consumers and end-users, and the governing instrument is the data security policy. "The policy reflects TP's commitment to protecting the privacy and personal data of each individual, including employees, suppliers, clients, business partners and end-users. The policy applies to all TP entities and ensures appropriate measures in processing data. It also constitutes a legal mechanism governing international data transfers within the Group, whenever TP acts as a data controller or processor, including when it transfers personal data on behalf of a client" (p.130).

The Group's dual role is set out first, because it determines where responsibility sits: "as data controller, it defines the purposes and means of processing personal data, mainly that of its employees; as data processor, it processes personal data on behalf of its clients, in accordance with contractual instructions" (p.129).

Operating principles include "Privacy by design and by default: collect only necessary data, delete in accordance with the storage schedule, no sale of personal data for marketing purposes", regular risk assessment of client processes, a procedure for data subject rights requests, and access to the Global Ethics Hotline (p.129).

Reference: pages 86, 130

Engagement is described in the terms that fit TP's position in the chain, which is that most end-users reach it as data subjects of a client's process. "The Group interacts with consumers and end-users, often in their capacity as data subjects, where they have requested access to their personal data or where regulators have submitted requests on their behalf. Satisfaction surveys are also sent to consumers to ascertain their opinions on the quality of customer service and the resolution of their requests" (p.130).

The stakeholder table treats end-users as a group in their own right (p.86): interests recorded as "Simple and rapid solutions to their daily requests; Privacy"; procedures as "Satisfaction surveys, omnichannel contacts"; purpose as "Improving end-user satisfaction"; the resulting changes as "Development of new services, privacy measures" and "Integration of innovative technologies"; and the route to directors as "Monitoring satisfaction rate and quality of services".

Not disclosed: the seniority of the person with operational responsibility for end-user engagement, how the effectiveness of engagement is assessed, and whether vulnerable end-users are engaged in any distinct way. The latter is consistent with the DMA, which found "protection of vulnerable users" not material (p.91).

Reference: pages 129-130, 134

The channel is the Group-wide hotline, applied to end-users: "A Global Ethics Hotline, accessible online, makes it possible to deal with the impacts and concerns of end-users confidentially and efficiently. All information on the hotline may be found in section 3.5.2.5 Global Ethics Hotline" (p.130). That section confirms the hotline is open to "any employee or stakeholder, internal or external" and covers human rights violations, discrimination, fraud and data matters, with alerts handled by a dedicated team attached to the Legal and Compliance Department and no retaliation tolerated (p.134).

Where responsibility sits is stated frankly, and it matters for what a reader should expect: "In the event of a breach of data security rules, responsibility lies with the client as the direct contact person for end-users. However, the Group may incur financial penalties, depending on contractual and regulatory obligations" (p.129).

Outcome data for the year is reported in the incidents table: 35 data privacy alerts, of which 17 were substantiated (p.134). The table does not separate end-user complaints from employee ones, and the report does not say whether end-users know the hotline exists or trust it.

Reference: pages 129-131

Actions are concentrated on the two things the DMA identified: data security and the ethical use of AI.

Certification and control framework (p.130): "The Group implements advanced strategies to ensure data security and cybersecurity, in compliance with international standards such as ISO 27701, ISO 27001, ISO 22301, GDPR, PCI-DSS and HITRUST. TP holds global ISO 27001 and ISO 27701 certifications." "Since 2018, the French Data Protection Authority (CNIL) has recognized the compliance of TP's Binding Corporate Rules (BCRs) authorizing the global transfer and processing of data." Internal and external audits verify subsidiary compliance, and "The privacy framework is based on proprietary technology designed to detect inappropriate access, render note-taking secure and ensure end-to-end compliance."

AI governance (p.131): six stated principles - responsibility with a "human-in-the-loop" approach, transparency, equity, privacy and security, reliability and security, and explainability - supported by an AI ethical use and governance policy and an AI evaluation policy. "TP is among the first companies being certified in accordance with ISO 42001 - AI management system", and "A committee dedicated to AI and innovation was created within the Board of Directors in 2025."

Reference: pages 87, 129

Two targets are set out in the material topic table at the head of the data privacy and cybersecurity section, each with a three-year series (p.129):

• "Percentage of employees trained in data security, privacy and data protection policies": 96% in 2023, 93% in 2024 and 87% in 2025, against a target of "> 90% each year". The company footnotes the fall: "As the training schedule has changed, this data only includes 9 months of the year, from April to December 2025", with the remaining employees required to complete training before 30 March 2026.
• "Percentage of eligible ISO 27001 and ISO 27701-certified facilities": 100% in each of 2023, 2024 and 2025, against a target of "100% each year".

The KPI dashboard on p.87 rates the training indicator as needing more action and the certification indicator as on target.

Both are input measures about TP's own control environment rather than outcome measures for consumers and end-users. There is no target for data privacy incidents, for complaint resolution time, or for end-user satisfaction with privacy, and the 35 data privacy alerts and 17 substantiated cases reported on p.134 are not set against any threshold.

The report does not say whether consumers and end-users, or their representatives, were involved in setting these targets or in tracking performance against them, and it gives no base year or target year: both are maintained annual thresholds.

Reference: pages 132-133

"As a pillar of TP's global strategy, corporate governance ensures ethical decision-making, effective internal controls and responsible conduct. It aims to ensure transparency, strengthen accountability and promote a culture of compliance and integrity" (p.132).

Code of Conduct and Ethics (p.133): "defines the standards and behaviors expected of all Group employees and entities, regardless of their status or function. Approved by senior management and available for consultation on the Group website, the Code applies to all stakeholder relations." Delivery relies on a network of Code ambassadors "raising awareness and supporting employees, providing practical advice and supporting management". Training reached 85% of employees between 1 April and 31 December 2025, against a target of over 90% each year, with the remainder due by 30 March 2026 (pp.132-133).

Culture and speak-up: the Global Ethics Hotline with guaranteed confidentiality and impartiality, and whistleblower protection integrated into the Global Ethics Hotline Policy (pp.133-134). The Appendix B table maps the UN Convention against Corruption datapoint (paragraph 10(b)) and the protection of whistleblowers datapoint (paragraph 10(d)) to section 3.5.2 (p.140). The Compliance Department "attached to the Board of Directors ensures that codes, policies and procedures are applied Group-wide" (p.132).

Reference: page 133

"TP is strongly committed to preventing and combating corruption and influence peddling, in accordance with key international and local legislation such as the Foreign Corrupt Practices Act, the Bribery Act and the French Sapin II Law" (p.133). The material IRO records the exposure it addresses: practices in conflict with anti-corruption, business ethics and tax evasion regulations arising in countries where the Group operates or in its value chain, rated at the highest criticality on the three-point scale (p.95).

Prevention, detection and response. "The program includes a regularly updated corruption risk map, a code of conduct defining expected behaviors and applicable sanctions, and a mandatory training program for all employees, supplemented by specific modules for Most Exposed Persons. A due diligence procedure is systematically applied prior to the signing of contracts with third parties... Accounting checks and internal audits are carried out to prevent and detect any acts of corruption, while performance and compliance indicators are assessed at least once a year."

Targeted training. "In 2025, 1,090 MEPs were identified, including senior management and key functions (procurement, legal, finance, audit, HR, sales, operations). As of 31 December, 761 employees had completed their training" - a 70% completion rate among the most exposed population.

Reference: pages 87, 132

Business conduct targets are reported through the ethics and anti-corruption material topic table, which pairs each KPI with a target and a three-year series (p.132):

• "Percentage of employees trained in the Code of Conduct": 95% in 2023, 89% in 2024 and 85% in 2025, against a target of "> 90% each year". The company footnotes the change of basis: "As the training schedule has changed, this data only includes 9 months of the year, from April to December 2025", and "The remaining employees must complete the training before March 30, 2026" (p.133).
• "Deployment of the Global Ethics Hotline at Group subsidiaries": 100% in each of 2023, 2024 and 2025, against a target of "100% each year".

The Group KPI dashboard repeats both under the Governance heading and rates progress against each, marking the Code of Conduct training indicator as needing more action and the hotline deployment as on target (p.87).

Both targets are maintained annual thresholds rather than dated trajectories: no base year, no target year and no interim milestones are given, and no target addresses outcomes such as substantiated incidents, of which there were 3 corruption cases in 2025 (p.134).

Reference: page 134

Incidents are reported in the Global Ethics Hotline table, which gives alerts received, alerts substantiated and the associated financial penalty (p.134):

• Corruption incidents: 13 alerts received, 3 substantiated, and "Amount of fines, sanctions and compensation related to cases of corruption or serious human rights violations (in euros): 0".
• Serious human rights incidents: 0 alerts, 0 substantiated, EUR 0.

For scale, the table sits within 3,340 reports received in 2025, of which 2,450 fell within the hotline's scope, against 1,643 reports in 2024.

Response is described at programme level: "In the event of a breach, disciplinary sanctions and corrective measures are implemented in accordance with local laws" (p.133), and "Corrective measures may include disciplinary sanctions, process improvements and targeted audits", with "Regular reports... presented to the Audit, Risk and Compliance Committee, while major cases are forwarded to senior management" (p.134).

Not disclosed: whether any of the 3 substantiated corruption cases involved the Group's own workers or business partners, the nature of the sanctions applied, and whether any convictions were recorded. The report gives the fine amount as zero but does not state the number of convictions separately.