freenet AG

Reference: pages 73-75 (ESRS 2 GOV-1, also listed for G1 in the index, page 143).

freenet AG has a dual board system: an Executive Board that manages the company and a twelve-member Supervisory Board (six shareholder and six employee representatives) that supervises it. On 1 September 2025 the Executive Board shrank from six members to two, both men, so the average female-to-male ratio for September to December was 0.0 (until 31 August 2025: 0.2). Five women sat on the Supervisory Board (weighted average ratio 0.7). All six shareholder representatives (100%) were independent under the German Corporate Governance Code.

ESG responsibilities (pages 73-75):

• The whole Executive Board is responsible at the operational level for sustainability-related impacts, risks and opportunities. Since 1 September 2025 the CFO has been responsible for the ESG strategy (previously the CHRO), and ESG Management moved to the Finance department.
• The Supervisory Board as a whole monitors ESG activities. The CFO reports to it, for example on the results of the climate risk and opportunity analysis.
• An interdisciplinary ESG Committee coordinates sustainability work across divisions. ESG Controlling has reported ESG key figures to the Executive Board quarterly since the start of 2025.
• Supervisory Board members self-assess their sustainability expertise in a skills matrix, and two members act as sustainability officers.

Reference: pages 75-76.

The ESG Committee is briefed by the ESG department during the annual update of the double materiality assessment (DMA) and at its monthly meetings on the status of policies, actions and targets. The CFO is responsible for implementing the DMA, and the entire Executive Board discusses, validates and decides on the results. The audit committee is informed annually about the audit of non-financial reporting and about changes in the materiality of sustainability matters.

The Chief Compliance Officer, who is also the Human Rights Officer, reports on human rights-related risks and remedial actions at the monthly ESG Committee meetings and at least once a year at a board meeting (page 76).

In 2025 the Executive Board, Supervisory Board and audit committee dealt with: adjusting the ESG organizational structure after the Executive Board was reduced; updating the prior-year materiality analysis; the results of the climate risk and opportunity analysis; progress on expanding the non-financial internal control system (N-ICS); quarterly monitoring of selected ESG key figures; and setting 2025 corporate targets with ESG components (page 76).

Reference: pages 76-77.

Executive Board pay has a fixed part plus a short-term incentive (STI) and a long-term incentive (LTI). Sustainability targets in use for 2025:

• STIP 2025: the rate of annual performance reviews, weighted 10%, tied to quantitative thresholds (see S1-13).
• LTIP 2025-2028: raising the device return rate to promote the circular economy (see E5-3).
• LTIP 2023-2026: employee satisfaction.
• LTIP 2022-2025: a Scope 1 and Scope 2 reduction target. The E1-4 target was formulated later and "it should be considered independently of this LTIP target".

Long-term sustainability targets in the LTIP are generally weighted at 20%. The Supervisory Board remuneration system "does not include any incentive system incorporating sustainability elements". The relative weighting and the amounts paid are set out in the remuneration report, not in this statement.

Reference: page 77.

freenet presents its due diligence statement as a table that maps the core elements of due diligence to ESRS disclosures:

• Embedding due diligence in governance, strategy and business model: ESRS 2 GOV-2, GOV-3 and SBM-3.
• Engaging with affected stakeholders: ESRS 2 GOV-2, SBM-2 and IRO-1, the topical ESRS and MDR-P.
• Identifying and assessing adverse impacts: ESRS 2 IRO-1 and SBM-3.
• Taking actions to address adverse impacts: ESRS E1-1, the topical ESRS and MDR-A.
• Tracking effectiveness and communicating: the topical ESRS, MDR-M and MDR-T.

Human rights due diligence under the German Supply Chain Due Diligence Act (LkSG) is described in GOV-2 (page 76), S1-1 (page 115) and S2 (pages 132-134).

Reference: page 78.

Sustainability risks and risks in sustainability reporting sit inside the group-wide risk management system (RMS) and internal control system (ICS), which follow COSO and IDW frameworks. The reporting-specific control system (N-ICS) "is continuously being developed". Departments collect the data. The ESG department, ESG Controlling and the ESG Committee quality-assure and approve material information.

Non-financial risks and opportunities are rated qualitatively, both gross and net of the effect of measures, using the same assessment ranges as general risk management. "No material risks were identified in relation to the sustainability reporting process."

The Group Audit department independently reviews the adequacy and effectiveness of the financial and non-financial control and risk systems and reports to the Executive Board and, regularly, to the audit committee of the Supervisory Board.

Reference: pages 79-80.

freenet AG calls itself a digital lifestyle provider. It offers mobile communications and TV services plus mobile devices and accessories, mainly to consumers in Germany. Its 3,085 employees are almost exclusively in Germany, so there is no geographic breakdown. The segments are Mobile Communications, TV and Media, and Other/holding. freenet "does not operate its own mobile network" and instead buys tariffs wholesale from German network operators for resale. In TV and Media it is a platform provider for internet-based TV (IPTV), terrestrial TV (DVBT) and digital terrestrial radio (DAB+). No products, services or markets changed in 2025.

Value chain: upstream is raw material extraction and processing, component manufacturing and final assembly of devices, with ICT device makers as direct suppliers. Downstream is logistics, franchises and customer service.

ESG strategy: fields of action are climate, circular economy, employees, customers and governance. freenet intends to reach the 1.5-degree path for Scope 1 and Scope 2 by 2030; Scope 3 is not yet in the transition plan (page 80).

Reference: pages 81-83.

freenet splits stakeholders into users of sustainability reporting (investors, lenders, analysts, rating agencies, NGOs) and affected stakeholders: own workforce, value chain workers, suppliers, customers and "nature as a silent stakeholder". A table (page 81) lists the groups and the engagement formats, such as investor meetings, works council and works meetings, employee surveys, town halls, the whistleblower system and customer surveys.

• Investors and analysts: dialogue centers on governance (including IT/cyber security) and on linking sustainability to Executive Board incentives. freenet reports climate data to CDP.
• NGOs: "There is no direct exchange" with them; their interests are taken into account through published information and studies.
• Value chain workers and nature: no regular dialogue with workers of companies without a direct contract; their views entered the DMA indirectly through public information and industry reports.
• DMA update: stakeholders were surveyed online in 2025, and the results went to the Executive Board and Supervisory Board (page 83).

Reference: pages 83-87.

The updated 2025 DMA confirmed, as in the prior year, material impacts in climate change (E1), circular economy (E5), own workforce (S1), workers in the value chain (S2) and business conduct (G1), plus an entity-specific topic, IT/cyber security. The table of material impacts and risks has 17 rows:

• E1: two actual negative impacts (value chain GHG emissions; high electricity use at Media Broadcast GmbH and in the shop chain) and one risk (rising carbon prices).
• E5: one actual negative impact (improper recycling or recovery of devices, downstream).
• S1: six potential and two actual negative impacts (working conditions, overtime, work-life balance, unequal pay, development, harassment, diversity awareness, one discrimination case).
• S2: two potential negative impacts (poor working conditions; child or forced labor in distant upstream stages).
• G1: one potential negative impact (bribery and corruption).
• IT/cyber security: one potential negative impact and one risk.

"As in the prior year, freenet has not identified any material positive impacts or material opportunities." No impact on strategy and business model was found, and no current financial effects were identified (pages 86-87). Climate resilience is also presented under E1-3 (2025 ESRS numbering).

Reference: pages 88-93.

freenet carried out its first DMA for FY2024 and updates it annually. The 2025 update checked topics against external CSRD practice analyses and a telecom peer analysis, considered M&A changes, surveyed key stakeholders and compared risks with SASB sector ESG risks. It followed EFRAG's Implementation Guidance 1, drew on ENCORE and other databases, and covered own operations plus the upstream and downstream value chain.

Scoring (page 90): impacts are scored 1 to 5 on scale, scope and, for negative impacts, irremediability. An actual impact is material when the three scores add up to 8 or more. Potential impacts combine average severity with a likelihood matrix, and for human rights severity takes priority over likelihood. Risks and opportunities are scored 1 to 5 on likelihood and financial magnitude, and a result of 3 or more is financially material, matching the "high" category in freenet's enterprise risk management.

The 2025 update "did not reveal any new material impacts, risks or opportunities". The ESG Committee approved the results and the Executive Board adopted them; the audit committee and works council were informed. Topic-specific IRO-1 sections for E1, E2, E3, E4, E5 and G1 follow (pages 91-93). Climate-specific risk identification and scenario analysis is also presented under E1-2 (2025 ESRS numbering).

Reference: pages 93-94 and 143-149 (appendix).

freenet followed the EFRAG guidance and mapped each material impact and risk to data points in EFRAG's data point list (Implementation Guidance 3). The appendix to the non-financial group statement holds the "List of disclosure requirements in ESRS covered by the sustainability report" (pages 143-146) and a list of data points from other EU legislation with a materiality column (pages 147-149).

What the index lists: ESRS 2 (BP-1, BP-2, GOV-1 to GOV-5, SBM-1 to SBM-3, IRO-1, IRO-2); E1-1 to E1-8; E5-1, E5-2, E5-3 and E5-5 (as an entity-specific key figure); S1-1 to S1-7, S1-9, S1-11, S1-13, S1-15, S1-16, S1-17; S2-1 to S2-5; G1-1, G1-3, G1-4; and the entity-specific IT/cyber security topic. E2, E3, E4, S3 and S4 data points carry "not material" labels.

For IT/cyber security freenet "based its qualitative disclosures on the structure of the topic-specific ESRS S4 standard" and its key figures on the G1 standard. With a few exceptions it omitted voluntary disclosures, and those it kept are marked as ESRS 1.114 disclosures.

Reference: pages 94 and 80.

freenet formulated its first transition plan for climate change mitigation in 2024, covering 2022 to 2030. It aims to cut Scope 1 and Scope 2 emissions by more than 70% by 2030 against the 2022 base year and is described as "consistent with the 1.5-degree target of the Paris Agreement" for those two scopes. The plan is "not comprehensive": Scope 3 emissions, first measured in 2024, are not included while freenet works out decarbonization levers.

Two levers carry the plan: (1) electrification of the vehicle fleet and (2) conversion of further locations to renewable energy, delivered through three actions (see E1-3). Implementation "does not require any significant investments" and can be funded from own means; freenet has no Taxonomy-aligned CapEx or CapEx plans.

Locked-in emissions sit in the heat supply of offices and shops and are assumed not to fall by 2030; "No transition risks could be identified" in connection with them. freenet is not excluded from the EU Paris-aligned Benchmarks. The Executive Board approved the plan in 2024 and the Supervisory Board discussed it.

Reference: pages 87 and 91.

Back-filled from ESRS 2 SBM-3 (page 87) and the E1 section of IRO-1 (page 91), where this content is disclosed in the FY2025 report. This disclosure requirement did not exist under the 2023 ESRS the report was prepared against.

In the second half of 2025 freenet ran a climate risk and opportunity analysis "in accordance with CSRD/ESRS" and stated for each material climate risk whether it is physical or transition (page 87). It used scenario analysis over three horizons, short term to 2030, medium term to 2040 and long term to 2050, after screening which assets and activities could be exposed (page 91).

• Physical risk: IPCC projections, focused on the high-emission scenario SSP5-8.5. A site analysis using geocoordinates compares projections with hazard limits, then tests vulnerability. "No significant physical climate hazards were identified for the 2025 analysis."
• Transition risk: a 1.5-degree scenario using the IEA Net Zero Emissions Scenario (World Energy Outlook 2024). It found no assets or activities incompatible with a climate-neutral economy.

The scope covered all business activities and the value chain, with no significant risk excluded. The text gives no global average temperature projection for either scenario. The material gross transition risk is rising CO2 prices.

Reference: page 87.

Back-filled from ESRS 2 SBM-3 ("Resilience of strategy and business model"), page 87, where this content is disclosed in the FY2025 report. This disclosure requirement did not exist under the 2023 ESRS the report was prepared against.

freenet ran a qualitative resilience analysis as part of its second-half 2025 climate risk and opportunity analysis. It covered all business activities and the upstream and downstream value chain, and its critical assumptions concerned CO2 pricing, law and politics, technology and the market. Short-, medium- and long-term periods run to 2030, 2040 and 2050. Mitigation actions and resources (E1-3) fed a separate net risk assessment.

Result: "Overall, freenet's business model appears to be resilient to physical and transitional climate risks." The one material gross transition risk, higher costs from rising CO2 prices, can be "significantly reduced" by the continued switch to renewable energy and electrification of the vehicle fleet. Insurance and building energy improvements address physical hazards.

The analysis "takes into account future developments that are inherently uncertain (e.g. the development of Scope 3 emissions)". No risky assets or business activities were identified, and access to finance at reasonable capital costs is not expected to be affected.

Reference: page 95.

freenet's climate policy is the E-Car policy, introduced in 2024 and applied in 2025. It regulates the procurement, purchase requirements and use of company cars and vehicles, with the aim of electrifying the vehicle fleet to cut the climate impact of combustion-engine vehicles. It covers mitigation only: "The areas of adaptation to climate change, energy efficiency and the use of renewable energies are not covered by the policy."

It applies to all permanent employees entitled to a company car and to pool and commercial vehicles. The whole Executive Board adopted it, and since the September 2025 restructuring the CFO is primarily responsible. The policy is on the intranet, and company guidelines are checked in internal audit reviews. Employees who order vehicles helped develop it and were told of the final version by email.

Reference: pages 95-96.

Three actions are decarbonization levers of the transition plan; two others are not.

• Fleet electrification (lever 1, Scope 1): about 23% electrified by end of 2025, cutting vehicle emissions by 299 t CO2e (prior year: 144 t). Planned effect by 2030: about 97% (2,075.8 t CO2e). A small part of the Media Broadcast service fleet cannot be converted because no suitable electric vehicles exist.
• Renewable electricity (lever 2a): 97.3% of electricity was already renewable in the 2022 base year. Converting the remaining sites would remove 1,208.3 t CO2e by 2030; 1,038 t was achieved by end of 2025 (prior year: 1,015 t).
• Nauen heating (lever 2b): replacing the oil heating by end of 2028 at the latest, about 114 t CO2e by 2030; 46 t cut so far through maintenance.
• Not levers: a 99 kWp solar plant at Büdelsdorf (the site was already on renewable power) and an ISO 50001 energy management system for Media Broadcast, with about 75% of sensors installed and certification scheduled for February 2026.

"No significant CapEx and/or OpEx are required" and all actions use internal funds.

Reference: pages 96-98.

The Scope 1 and Scope 2 target, set in 2024, is a cut of about 73% by 2030 against 2022. The base year is 4,663.7 t CO2e (adjusted for the discontinued Gravis business, the sold data center and The Cloud Networks); the 2030 target is 1,266.1 t (-72.9%): Scope 1 at 1,051.7 t (-67.6%) and Scope 2 at 214.4 t (-84.9%). For comparison, the SBTi cross-sector 1.5-degree path (-42%) gives 2,705.0 t. These are gross targets with no removals or CO2 certificates, and Scope 2 uses the market-based method.

There is no Scope 3 target; those emissions "are still undergoing internal evaluation". The target was set internally and "was not externally validated".

Progress: the CFO compares actual emissions with the planned annual reduction. In 2025, emissions adjusted for events fell 31% (1,431 t CO2e) against the prior year (prior year: 26%, 1,232 t). Emissions covered by the plan were 3,232.9 t CO2e, "in line with the planned progress".

Reference: pages 99-100.

Total energy consumption was 78,927.9 MWh (2024: 85,157.0 MWh, down 7.3%; 2022 base year: 87,659.9 MWh). Fossil sources were 14,757.9 MWh, or 18.7% of the total (2024: 19.5%), and renewable sources 81.3% (2024: 80.5%). Energy covers vehicle fuel, electricity and heating.

The photovoltaic plant at Büdelsdorf produced 84.3 MWh, of which 80.4 MWh was consumed on site and counted as self-generated renewable energy.

Method: fuel use comes from fuel cards (100% primary data). Electricity and heat use a hybrid of recorded and projected consumption. For shops, consumption is average use per square meter times leased area, and the primary data share for that average is 37.5%. Fuel consumption from biomass is not reported "Due to limited data availability".

Reference: pages 100-106.

The largest Scope 3 categories are purchased goods and services (30,679.8 t, down 54%) and use of sold products (20,211.3 t, down 47%). GHG intensity per EUR million of revenue is 30.6 t market-based and 40.1 t location-based (2024: 54.3 and 64.7).

Category 10 is excluded because freenet's downstream chain has no such activity. Prior-year figures were restated for improved data, and categories 3.1, 3.4, 3.11 and 3.12 were re-estimated. Scope 3.11 is flagged as subject to high measurement uncertainty (page 71). Category 15 counts pro rata Scope 1 and Scope 2 of the Ceconomy AG and Antenne Deutschland stakes. All purchased electricity and heat carries bundled renewable certificates (100%).

Reference: page 107.

freenet reports a nil position. "freenet AG does not engage in removal of greenhouse gases or greenhouse gas reduction projects financed by CO₂ certificates." This matches the climate target in E1-4, which the company describes as a gross target: "neither removal of greenhouse gases nor CO₂ certificates are assumed" (page 97). The E1-7 line appears in freenet's ESRS index (page 144), and the Scope 1 and Scope 2 reduction path rests on fleet electrification and renewable energy conversion alone. The report also states that no biogenic CO2 emissions arose in the reporting year (page 101).

Reference: page 107.

freenet reports a nil position: "freenet AG does not apply an internal CO₂ pricing scheme." The E1-8 line appears in freenet's ESRS index (page 144). Carbon prices still matter to the company's risk picture. Rising prices for greenhouse gas emissions caused by stricter regulation are one of the material climate risks (page 84), and the resilience analysis took critical assumptions on CO2 pricing into account (page 87). freenet's report ties this risk to higher procurement prices for ICT equipment and accessories and higher logistics costs (page 85).

Reference: page 107.

freenet reports that it has no circular economy policy and explains why. Its model centers on selling mobile devices, mainly smartphones, with tariffs. After a device change the old phone often stays unused with the customer or is not disposed of properly. freenet offers a trade-in program online and in shops, run with a specialized service provider where the recycling and treatment take place.

Returning a device is voluntary for customers, which freenet can encourage but "cannot prescribe or formalise". Given its "purely intermediary role", the company considers a circular economy policy for mobile devices not appropriate at this time, because an internal policy would have no direct effect on the behavior of customers and other outside actors and so "would therefore have no substantial impact".

Reference: pages 107-108.

freenet's main action is a trade-in service in shops and the online shop, aimed at getting unused devices back into reuse, recycling or proper disposal.

• Incentives: with device manufacturers, "bonus promotions" add a bonus on the purchase price of a new device when a customer returns an old one. In 2025 freenet added a billing option that credits the returned device's value to the customer account, and promoted trade-in through newsletters and flyer inserts.
• Partner: the service runs with DIS-CONNECT under an open-ended contract. In 2025 a new tool for accepting devices in shops was introduced and the sales staff training portal was revised.
• Disposal-only option: customers can print a shipping label on the freenet brand website to send in devices for proper disposal, without compensation.

The IT changes and fewer bonus campaigns "led to a decline in the use of the purchase service" in 2025. The actions ran all year, are to continue, and freenet is evaluating their effectiveness.

Reference: page 108.

For the circular economy field of action freenet set a voluntary target at the start of 2025: raise the return rate to 6%. The return rate is the ratio of returned to sold devices (see E5-5). It is a four-year target up to and including 2028 with no interim targets, and it covers take-back of end-of-life devices, online and in shops, in the mobile communications segment. The target does not refer to a level of the waste hierarchy: it counts all devices taken back, whether reused, recycled or disposed of.

The target is part of the Executive Board's long-term variable remuneration (LTIP), set by the Supervisory Board. It is not linked to a company policy, because none exists for this field, and "No scientific findings, methods or significant assumptions were used in defining the target." Progress: the 2025 return rate was 1.4% (prior year: 3.9%).

Reference: page 108 (index entry, page 144).

freenet's ESRS index lists E5-5 as an "Entity-specific key figure" for the return rate of devices, not as the ESRS resource outflow datapoints. The company's own data point table marks the two E5-5 waste datapoints (non-recycled waste, hazardous waste) as "not material" (page 148). The key figure is the ratio of mobile devices accepted through freenet's purchase and recycling programs to the three-year average of mobile devices sold.

The three-year average denominator avoids swings from yearly sales changes. The main limitation, in freenet's words, is that the figure gives no information about the use of devices after hand-over to the service provider. Inputs for devices sold and accepted rely on 100% primary data.

Reference: pages 115-117.

freenet manages its material workforce impacts through six policies: the policy statement on the protection of human rights, the learning culture and principles for continuous skills development, the commitment to changing life realities, the competence model and annual performance review, the principles of equal pay, and the anti-discrimination policy. The human rights statement follows the UN Universal Declaration, the UN Global Compact, the UN Guiding Principles, the ILO Declaration on Fundamental Principles and Rights at Work and the German Supply Chain Due Diligence Act, and covers employees, non-employees and the supply chain. It addresses all material workforce topics except work-life balance, which the commitment to changing life realities covers.

Responsibility sits with the Executive Board or the CEO, working with the HR department. On discrimination, freenet states it has no specific policies beyond the equal pay principles and the legal framework (Works Constitution Act, AGG). On secure employment it refers to German social legislation rather than a company policy. Human rights concerns can be raised through the whistleblower system.

Reference: pages 119-122.

freenet groups its workforce actions into four areas.

• Work-life balance and working time: part-time work, flexible shift models in logistics, home office and mobile work, a hybrid model with three office days a week, health days, occupational health examinations and stress training. Overtime accounts are reviewed twice a year. Quantifiable results are "not currently being sought".
• Anti-discrimination and diversity: signed the Diversity Charter in 2024, diversity training (including for HR managers in Q1 2025), and coaching for women.
• Skills development: learning management systems for different employee groups, onboarding for new managers and annual development talks.
• Gender equality and equal pay: a pay analysis and a roadmap to close unjustified pay gaps by the end of 2026 at the latest.

Budgets exist for each area but are "not considered significant in relation to the overall budget". No actions to achieve positive impacts on the workforce were implemented.

Reference: page 122.

freenet has one numeric workforce target, set by the Executive Board in early 2022: women in leadership by 31 December 2026.

The tier 2 target is below the 2021 level because known fluctuation was built in when it was set. Group HR Controlling collects the quota quarterly. A second target is a 95.6% annual performance review rate for eligible employees (counts as 100% achieved), which drives the bonus (see S1-13 and GOV-3). Both targets were set and are tracked without direct stakeholder engagement, and none changed in 2025.

Reference: pages 123-125.

freenet reports the average number of employees: 3,085 in 2025 (2024: 3,450), of whom 915 are women, 2,170 men and 1 not specified. The total includes 156 vocational trainees, 21 dual students and 3 interns, whom freenet includes voluntarily.

Employee turnover was 369 people or 12.9% (2024: 809, 25.4%). In shops it was 20.0% (51.2%), mostly due to the wind-down of Gravis in 2024, and 10.5% (15.4%) excluding shops. There is no country breakdown because freenet has no country other than Germany with 50 or more employees or 10% of its workforce. Data come mainly from the HR system, with manual consolidation for some companies.

Reference: page 125.

freenet had an average of 22 non-employees in 2025 (2024: 28). As in the prior year, all were temporary agency workers, mostly in logistics and on-site IT support. The company counts non-employees as self-employed people who contract with it for their labor and people supplied by staffing firms. The average is built from month-end numbers divided by 12, mainly from the HR system, with manual data from some companies and requests to external staffing providers. "There are no estimation uncertainties." Non-employees are not included in the employee figures in S1-6. The SBM-3 text notes that non-employees are covered by the same material impacts as employees (page 85).

Reference: pages 125-126.

Gender at the top two management tiers (first and second tier below the Executive Board): an average of 119 people in 2025 (2024: 131), of whom 31 were women (25.7%; 2024: 36, 27.3%) and 89 men (74.3%; 2024: 95, 72.7%). None were recorded as other or not specified.

Age distribution of employees (3,085):

• Under 30: 678 (22.0%; 2024: 23.1%)
• 30 to 50: 1,487 (48.2%; 2024: 48.0%)
• Over 50: 920 (29.8%; 2024: 28.9%)

Figures use monthly averages, the HR system and manual consolidation for companies not in that system. freenet states "There are no estimation uncertainties." Executive and Supervisory Board gender ratios are in GOV-1 (page 73).

Reference: page 126.

In 2025 all employees (prior year: 100%) were covered against loss of earnings from major life events by the social security systems of the Federal Republic of Germany. The statutory framework covers illness (health and long-term care insurance), unemployment, accidents at work and incapacity to work (accident insurance), parental leave and retirement (pension insurance). Beyond the legal minimum, freenet gives employees subsidies for occupational pension schemes, and since 2016 it has offered occupational disability insurance with a simplified health check.

Reference: page 127.

The review figure follows the ESRS definition and matches the competence model and annual performance review in S1-1. Training hours cover documented formal learning, using learning management system data and manual reports from HR managers. They exclude hours at vocational schools and universities, compulsory training and paid training leave. For platforms with permanent access and no exact time recording, hours rest on partner estimates of average useful life. The calculation note also cites 83.3% (prior year: 71.5%) next to the average number of eligible employees.

Reference: page 128.

All employees in the Group have a statutory right to leave from work for family reasons under German law, including maternity, paternity and parental leave and leave to care for relatives.

The share is the number of employees who took such leave divided by the total (by gender category), using HR system data and manual consolidation for some companies. The main reasons were maternity leave and parental leave. freenet states there are "no estimation uncertainties".

Reference: pages 128-129.

The gender pay gap is the difference between men's and women's average standardized annual target salary (fixed plus variable at 100% target achievement, projected to full-time), divided by men's average. The gap "is primarily due to the fact that at freenet, more men than women hold managerial or specialist positions". The pay ratio compares the highest-paid person, the CEO as in the prior year, with the median of all employees, excluding Executive Board members and managing directors. Long-term incentive tranches are counted in the year of allocation on the recorded expense. Median calculation uses approximations based on income subject to tax.

Reference: pages 129-130.

For the one discrimination incident freenet launched an investigation as soon as it learned of it, suspended the manager concerned and terminated their employment. A manager who failed to fulfil their responsibilities was given a warning and required to attend prevention seminars. There were no confirmed serious human rights violations among the own workforce. Discrimination is defined by the grounds in the AGG, and the figures are collected manually by Group HR Controlling through a survey of managers and the reporting channels.

Reference: pages 130-131.

freenet and its affiliated companies committed in 2023 to protecting internationally recognized human rights in a joint policy statement on the company website. Suppliers and business partners must follow the group-wide Supplier Code of Conduct, which sets minimum standards on human rights (zero tolerance for human trafficking, forced labor and child labor), social standards such as legal wages, environmental protection, safety, health and compliance. The code is included in all new procurement contracts, or strategic suppliers declare standards at least equivalent. Commitment to the code and to the UN Global Compact, UN Guiding Principles and ILO Declaration is also built into the General Terms and Conditions of Purchase.

On violations freenet "reserves the right to take appropriate action", up to ending the contract. The Chief Compliance Officer is the Human Rights Officer. freenet says its ability to influence its main suppliers is limited by its share of their business volume. It knew of no case in 2025 (prior year: 0) of a breach of the UN Guiding Principles, the ILO Declaration or the OECD Guidelines involving value chain workers (page 131).

Reference: pages 133-135.

freenet identified human rights violations and working conditions in the value chain as material. Its actions rest on the assumption that most of its sales volume comes from partners in EU member states, the EEA or OECD countries with comparable legal requirements. The initial LkSG risk analysis "did not reveal any specific material human rights risks" for its own business or suppliers.

• The Supplier Code of Conduct is in all new procurement contracts; breaches can bring sanctions up to termination.
• All direct suppliers in scope were surveyed in 2023 with LkSG questionnaires. New direct suppliers are questioned on a risk basis during setup, and Compliance reviews the answers. Remediation follows Section 7 LkSG and can go as far as ending the relationship.
• A whistleblower system is available around the clock, anonymously or with identity.

freenet "does not currently plan any additional actions or initiatives specifically aimed at promoting positive impacts" on value chain workers. No serious human rights problems were reported in its upstream or downstream chain in 2025 (prior year: 0). LkSG risk management is part of Group Internal Audit's audit universe.

Reference: page 135.

freenet reports that it has set no targets for value chain workers. "freenet has not set any measurable, results-oriented targets for managing impacts on value chain workers and currently has no plans to do so." Its reasons: the business model and the immediate upstream and downstream chain are concentrated in Germany, and it assumes its influence on sustainability impacts for value chain workers is low, a view it says was confirmed by the LkSG risk analysis.

It monitors the effectiveness of its policies and actions through LkSG risk management. "Beyond this, the company has not defined sufficiently specific targets or agreed on qualitative or quantitative indicators." Its stated aim, in S2-1, is to hold suppliers, phone makers and network operators accountable for human rights due diligence and the exclusion of conflict minerals (page 130).

Reference: pages 139-140.

freenet describes its corporate culture as marked by integrity, cooperation and continuous learning, with a zero-tolerance policy on corruption. Corruption and bribery were identified as material in the DMA. The Executive Board and managers communicate the "tone from the top", and works councils support the compliance policies.

Three policies carry the anti-corruption work: the gift policy (all gifts received or given above EUR 20 are reported to compliance quarterly), the signatures policy (dual control, so only authorized people sign contracts and legally binding declarations) and the purchasing policy (the purchasing department takes part in material acquisitions as a neutral body). All sit inside the Compliance Management System (CMS), which also gives employees low-threshold legal advice. The CMS covers the whole value chain, and a code of conduct for business partners has been part of B2B contracts for many years. The Chief Compliance Officer owns the policy and reports to the CFO. Most exposed are purchasing, network operator management and sales.

Reference: pages 139-142.

Back-filled from the business conduct chapter, where targets and effectiveness tracking are handled under the minimum disclosure requirements rather than as a numbered disclosure requirement. G1-3 became a standalone DR only in the 2025/2026 ESRS.

The chapter states no numeric business conduct target. It tracks effectiveness instead:

• The Chief Compliance Officer reports at least once per financial year to the audit committee so it "can satisfy itself as to the effectiveness of the CMS", and the audit committee "monitors the effectiveness of this system" (page 141).
• Training coverage is measured: about 75% of employees and Executive Board members in exposed roles trained in 2025 (prior year: 44.8%), with more than 70% of employees in the most exposed areas through advanced training (page 141).
• Outcome counts are reported: no convictions and no fines in 2025 or 2024 (page 142).
• All policies are reviewed regularly for compliance with legal requirements and ethical developments (page 139).

Reference: page 142.

freenet is "not aware of any convictions or fines for violations of corruption and bribery regulations for the 2025 financial year" and not aware of any violations of procedures and standards on anti-corruption and anti-bribery. The convictions figure counts confirmed cases in the year that led to criminal sanctions against employees or members of executive bodies. The fines figure is the total paid by those people in the year for violations in that year or earlier, in euros. Neither figure carries limits or significant assumptions. The report states "The key figures have not been validated by an external body".