Report September 2025

Submitted

Your organisation description

Permanent Task-Force

Commitment 37

Signatories commit to participate in the permanent Task-force. The Task-force includes the Signatories of the Code and representatives from EDMO and ERGA. It is chaired by the European Commission, and includes representatives of the European External Action Service (EEAS). The Task-force can also invite relevant experts as observers to support its work. Decisions of the Task-force are made by consensus.

We signed up to the following measures of this commitment

Measure 37.1 Measure 37.2 Measure 37.3 Measure 37.4 Measure 37.5 Measure 37.6

In line with this commitment, did you deploy new implementation measures (e.g. changes to your terms of service, new tools, new policies, etc)?

If yes, list these implementation measures here

There have been no significant updates since the last submitted report.

Do you plan to put further implementation measures in place in the next 6 months to substantially improve the maturity of the implementation of this commitment?

If yes, which further implementation measures do you plan to put in place in the next 6 months?

Measure 37.6

Signatories agree to notify the rest of the Task-force when a Commitment or Measure would benefit from changes over time as their practices and approaches evolve, in view of technological, societal, market, and legislative developments. Having discussed the changes required, the Relevant Signatories will update their subscription document accordingly and report on the changes in their next report.

QRE 37.6.1

Signatories will describe how they engage in the work of the Task-force in the reporting period, including the sub-groups they engaged with.

There have been no significant updates since the last submitted report.

Monitoring of the Code

Commitment 38

The Signatories commit to dedicate adequate financial and human resources and put in place appropriate internal processes to ensure the implementation of their commitments under the Code.

We signed up to the following measures of this commitment

Measure 38.1

In line with this commitment, did you deploy new implementation measures (e.g. changes to your terms of service, new tools, new policies, etc)?

If yes, list these implementation measures here

Globally we have around 40,000 people working on safety and security.

Do you plan to put further implementation measures in place in the next 6 months to substantially improve the maturity of the implementation of this commitment?

If yes, which further implementation measures do you plan to put in place in the next 6 months?

As mentioned in our baseline report, our policies benefit from our experience and expertise.

Measure 38.1

Relevant Signatories will outline the teams and internal processes they have in place, per service, to comply with the Code in order to achieve full coverage across the Member States and the languages of the EU.

QRE 38.1.1

Globally we have around 40,000 people working on safety and security including around 15,000 content reviewers. All of these investments work to combat the spread of harmful content, including disinformation and misinformation, and thereby contribute to our implementation of the Code.

Teams with expertise in content moderation, operations, policy design, safety, market specialists, data and forensic analysis, stakeholder and partner engagement, threat investigation, cybersecurity and product development all work on these challenges. These teams are distributed globally, and draw from the local expertise of their team members and local partners.

Commitment 39

Signatories commit to provide to the European Commission, within 1 month after the end of the implementation period (6 months after this Code’s signature) the baseline reports as set out in the Preamble.

We signed up to the following measures of this commitment

In line with this commitment, did you deploy new implementation measures (e.g. changes to your terms of service, new tools, new policies, etc)?

If yes, list these implementation measures here

Do you plan to put further implementation measures in place in the next 6 months to substantially improve the maturity of the implementation of this commitment?

If yes, which further implementation measures do you plan to put in place in the next 6 months?

Crisis and Elections Response

Elections 2025

[Note: Signatories are requested to provide information relevant to their particular response to the threats and challenges they observed on their service(s). They ensure that the information below provides an accurate and complete report of their relevant actions. As operational responses to crisis/election situations can vary from service to service, an absence of information should not be considered a priori a shortfall in the way a particular service has responded. Impact metrics are accurate to the best of signatories’ abilities to measure them].

Threats observed or anticipated

Meta is committed to providing reliable election information while combating misinformation across languages on our platforms. Our policies and safeguards for elections have been developed over many years and informed by our experiences of working on more than 200 elections around the world. Those experiences have resulted in the development of a robust election program, which uses mature policies, processes, and tools to both protect speech on our platform and safeguard the integrity of the elections. We continuously improve these measures to ensure they remain appropriate and responsive to emerging risks, and we have reinforced these efforts in light of the regulatory framework set out under the Digital Services Act, the Election Guidelines, and our commitments under this Code.

Community Standards and Guidelines Relevant to Elections:

Our Community Standards set out strict rules for content that can and cannot be posted to our platforms. These policies cover voter interference, voter fraud, electoral violence, and misinformation, among other categories such as hateful conduct, coordinating harm and promoting crime, bullying and harassment. Our policies have been refined over many years, by partnering with academics, civil society, and third-party fact-checkers to find the appropriate balance between protecting people and protecting freedom of expression and information. These policies are regularly reviewed, and they are made available to the public through our Transparency Centre.

Our comprehensive approach to elections continued for European elections held between 1 January - 30 June 2025. The election responses covered in this report include:

Germany (Parliamentary), 23 February 2025
Romania (Presidential), 4 May, 2025
Romania (Presidential Runoff), 18 May 2025
Portugal (Parliamentary), 18 May 2025
Poland (Presidential), 18 May 2025

Mitigations in place

Our Election Risk Management Processes

We have a dedicated team responsible for driving Meta’s cross-company election integrity efforts, leveraging experts from a full range of business functions to foster a holistic approach to tackling election-related risks. Those functions include colleagues in Meta’s intelligence, data science, product and engineering, research, operations, content and public policy, and legal teams.

Building on our experience of the 2024 European Parliament (EP) Elections, we continued to conduct in depth preparations and risk assessments for elections covered in this reporting period, deploy mitigation measures and utilise our Election Operation Centres established to address risks in real time ahead of the elections day.

We continued to work closely with a full range of external stakeholders to inform our processes and procedures ahead of elections. This included collaboration with Member State Digital Service Coordinators (DSCs), national authorities, electoral bodies, as well as taking part in the EU Code of Practice (“CoP”) Rapid Response System. As part of the rapid response system framework, we onboarded designated civil society organisations and fact checkers to our direct escalation channels to report time sensitive content, accounts or trends that could threaten the integrity of the electoral process.

Overview of Cooperation with External Stakeholders and Election Integrity Efforts

Meta engages with a full range of external stakeholders to inform our processes and procedures as part of our day-to-day business, and this practice continued during our election preparation and integrity efforts for Germany, Romania, Portugal and Poland. Meta values the networks and channels we have with our external stakeholders to work together in identifying risks on our platforms, and as such, we have welcomed many of the Election Guidelines recommending cooperation and points of contact with national authorities, civil society organisations, and others.

Germany

External engagement and election preparations started in the second half of 2023 as part of Meta’s overall 2024 EU Parliamentary election integrity efforts. For the 2025 German Federal elections, these efforts included participating in over 15 engagements with German and EU Level authorities including: the Ministry of Interior (MoI), the German Digital Service Coordinator (DSC) “BNetzA”, the German Government Election Taskforce, the Federal Office for Information Security (BSI), the Federal Returning Officer and German intelligence services. Many of these engagements took the form of bilateral meetings or round tables chaired by the German MoI, or the BNetzA in close partnership with the European Commission. We onboarded the German DSC to our direct regulatory reporting channel, as well as other relevant authorities such as the Federal Returning Officer, the BSI or the For the German election taskforce.

Meta also participated in the 'DSA Stress Test: Tabletop Exercise on German Elections' organised by the German DSC, BNetzA, and the European Commission. The event brought together representatives from social media platforms, as well as national authorities and civil society organisations, who took part in the stress test exercise.

In close partnership with the BSI, Meta organised a capability building session reaching more than 90 MPs, members of Parliamentary Groups as well as candidates. The training focussed on Meta’s election preparedness narrative, our advertising and organic best practices, business messaging and general safety updates to raise awareness of security threats and possible misinformation and disinformation campaigns around the elections.

Ahead of the German Elections, we also partnered with the German Returning Officer to support their “Get out the vote” campaign through ad credits, reaching almost 18 million users and more than 26 million video plays.

Overview of partners and notifications received during the Rapid Response Implementation period (6 February to 5 March):

Number of onboarded non-platform signatories to our direct reporting channels: 6.
Number of reports received during the election period through the rapid response system: 18.

Voter Information Units and Election Day Information Features

We remain focused on providing users with reliable election information while combating misinformation across languages. That is why we continue to connect people with details about the election for their Member State through in-app notifications, where legally permitted. We proactively point users to reliable information on the electoral process through in-app ‘Voter Information Units (VIU)’ and ‘Election Day Information’ reminders (EDR).

FacebookVIU Reach: Over 17.4 millionEDR Reach: Over 11.3 million | InstagramVIU Reach: Over 29.5 millionEDR Reach: Over 23.4 million

Romania (First Round and Run Off)

The election preparations efforts for Romania started in the second half of 2023 as part of the overall 2024 EU Parliamentary elections preparation efforts, and continued until the 2025 Presidential election. Meta engaged with a full range of external stakeholders to inform our processes and procedures. This included regular engagement with Romania’s Permanent Electoral Authority (PEA), the Romanian Digital Service Coordinator “AnCOM,” the Ministry of Digitalisation, Research and Innovation, the Ministry of Interior, the Cyber Security Directorate (DNSC) and the Audiovisual Council, all of whom were onboarded to our direct regulatory escalation channels where they were able to report content.

As an active member of the EU Code of Practice on Disinformation Taskforce’s Working Group on Elections, we took part in its Rapid Response System. Through this, we were regularly in touch with civil society organisations from Romania through various meetings and roundtables organised by the Disinformation working group. In addition to this, we also participated in a Stress Test organised by AnCOM in Bucharest, alongside national authorities, civil society and other very large online platforms.

Meta also engaged Romanian Political Parties in advance by organising online training sessions on our policies and products, including how to contact Meta in case of an escalation. Meta also created a direct escalation channel for 5 Romanian partners to report Community Standards violations and unlawful content, and collaborated with the Electoral Body to support civic engagement for Romanian users and connect people with reliable information about voting.

Overview of partners and notifications received during the Rapid Response Implementation period: (7 April to 25 May)

Number of onboarded non-platform signatories to our direct reporting channels: 7.
Number of reports received during the election period: 60.

Voter Information Units and Election Day Information Features

FacebookFirst RoundEDR Reach: Over 6.9 millionRunoff EDR Reach: Over 7.2 million | Instagram First round:EDR Reach: Over 3.0 millionRunoff EDR Reach: Over 2.9 million

Portugal

Election preparations efforts for the 2025 Portugal election started in the first half of 2024, following the announcement of the snap election. This included establishing formal communication channels with the Portugal Electoral Commission (CNE), the Portuguese Digital Service Coordinator, “ANACOM,”and comprehensive outreach to each political party to ensure that political parties and candidates' teams were aware of the critical resources, policies, and escalation channels.

Meta also supported civic engagement for Portuguese users by collaborating with the Portuguese Electoral Body to connect people with reliable information about voting. In the lead up to the election and on the election day, Meta showed on top of feed notifications on both Facebook and Instagram to all users in Portugal to redirect them to their website.

Overview of partners and notifications received during the Rapid Response Implementation period (7 April to 25 May):

Number of onboarded non-platform signatories to our direct reporting channels: 2.
Number of reports received during the election period: 2.

Voter Information Units and Election Day Information Features

FacebookVIU Reach: Over 4.6 millionEDR Reach: Over 3.3 million | InstagramVIU Reach: Over 5.2 millionEDR Reach: Over 4.3 million

Poland

Meta conducted a series of targeted initiatives to enhance external collaboration with key stakeholders ahead of the Polish election, engaging close to 200 stakeholders from across government, politics, academia, and NGOs. This included engagements and workshops with: the National Electoral Office, representatives from the Cybersecurity Directorate, Internal Intelligence, Counter Espionage, and Police HQ to streamline cooperation in processing data requests and escalations. We also conducted engagements with the Polish Research and Academic Computer Network (NASK) and Ministry of Foreign Affairs to aid understanding of our content escalation channel and provide training on key content policies to improve moderation and reporting.

In addition to this, one-on-one workshops with teams from each registered presidential candidate were arranged by Meta. These sessions focused on clarifying our policies, establishing communication channels, and providing access to "Meta Support Pro" for priority technical issue resolution. Training was also provided to the Polish Ministries, including with the Chancellery of the Prime Minister, to train their communication teams on effective cybersecurity incident prevention.

As part of the rapid response system, Meta maintained regular contact with civil society organisations and created a direct escalation channel for Polish partners to report Community Standard violating and unlawful content. This included: Alliance4europe, CEE Digital Democracy Watch, GLOBSEC and DEMAGOG.

Overview of partners and notifications received during the Rapid Response Implementation period (22 April to 24 June 2025):

Number of onboarded non-platform signatories to our direct reporting channels: 4.
Number of reports received during the election period: 14.

Voter Information Units and Election Day Information Features

FacebookFirst RoundEDR Reach: Over 12.8 millionRunoff EDR Reach: Over 12.4 million | InstagramFirst RoundEDR Reach: Over 7.3 millionRunoff EDR Reach: Over 7.0 million

Responsible Approach to Gen AI

Meta’s approach to responsible AI is another way that we are safeguarding the integrity of elections globally, including for the EU national elections.

Community Standards, Fact-Checking, and AI Labelling:

Meta’s Community Standards and Advertising Standards apply to all content, including content generated by AI. AI-generated content is also eligible to be reviewed and rated by Meta’s third-party fact-checking partners, whose rating options allow them to address various ways in which media content may mislead people, including but not limited to media that is created or edited by AI.

Meta labels photorealistic images created using Meta AI, as well as AI-generated images from certain content creation tools.

Meta has begun labelling a wider range of video, audio, and image content when we detect industry-standard AI image indicators or when users disclose that they are uploading AI-generated content. Meta requires people to use this disclosure and label tool when they post organic content with a photorealistic video or realistic-sounding audio that was digitally created or altered, and may apply penalties if they fail to do so. If Meta determines that digitally created or altered image, video, or audio content creates a particularly high risk of materially deceiving the public on a matter of importance, we may add a more prominent label, so that people have more information and context.

Continuing to Foster AI Transparency through Industry Collaboration:

Meta has also been working with other companies in the tech industry on common standards and guidelines. Meta Platforms, Inc. is a member of the Partnership on AI, for example, and signed onto the tech accord designed to combat the spread of deceptive AI content in 2024 elections globally. Meta receives information from Meta Platforms, Inc. in the progress of these initiatives, and benefits from these partnerships when addressing the risks of manipulated media.

Scrutiny of Ads Placements

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

The measures outlined in Chapters 1 to 3 of this report were in place for the elections covered in this report. They were complemented by the prohibited ads policy outlined above. Most pertinently, under these policies, content that is fact-checked cannot be used for an ad under our Advertising Standards.

Political Advertising

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

As outlined in Section 6, Beginning in October 2025, Meta will no longer allow political, electoral and social issue ads on our platforms in the EU, given the unworkable requirements and legal uncertainties introduced by the EU’s Transparency and Targeting of Political Advertising regulation.

Integrity of Services

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

All the measures outlined in Chapters 14 to 16 of this report were in place ahead of the European national elections.

Empowering the Research Community

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

Since 2023, researchers in Europe have had access to the Meta Content Library, enabling them to study various topics, including disinformation.

Crisis 2025

Threats observed or anticipated

War of aggression by Russia on Ukraine

As outlined in our benchmark report, we took a variety of actions with the objectives of:

Helping to keep people in Ukraine and Russia safe: We’ve added several privacy and safety features to help people in Ukraine and Russia protect their accounts from being targeted.
Enforcing our policies: We are taking additional steps to enforce our Community Standards, not only in Ukraine and Russia but also in other countries globally where content may be shared.
Reducing the spread of misinformation: We took steps to fight the spread of misinformation on our services and consulted with outside experts.
Transparency around state-controlled media: We have been working hard to tackle disinformation from Russia coming from state-controlled media. Since March 2022, we have been globally demoting content from Facebook Pages and Instagram accounts from Russian state-controlled media outlets and making them harder to find across our platforms. In addition to demoting, labelling, demonetizing and blocking ads from Russian State Controlled Media, we are also demoting and labelling any posts from users that contain links to Russian State Controlled Media websites.
In addition to these global actions, in Ukraine, the EU and UK, we have restricted access to Russia Today (globally), Sputnik, NTV/NTV Mir, Rossiya 1, REN TV and Perviy Kanal and others.
On 15 June 2024, we added restrictions to further state-controlled media organisations targeted by the EU broadcast ban under Article 2f of Regulation 833/2014. These included: Voice of Europe, RIA Novosti, Izvestia, Rossiyskaya Gazeta.
On 17 September 2024, we expanded our ongoing enforcement against Russian state media outlets. Rossiya Segodnya, RT, and other related entities were banned from our apps globally due to foreign interference activities.

[Israel - Hamas War]

In the spirit of transparency and cooperation we share below the details of some of the specific steps we are taking to respond to the Israel - Hamas War.

Mitigations in place

[War of Aggression by Russia on Ukraine]

Our main strategies are in line with what we outlined in our benchmark report, with a focus on safety features in Ukraine and Russia, extensive steps to fight the spread of misinformation (including through media literacy campaigns), tools to help our community access crucial resources, transparency around state controlled media and monitoring/taking action against any coordinated inauthentic behaviour.

This means (as outlined in previous reports) we will continue to:

Monitor for coordinated inauthentic behaviour and other adversarial networks (see commitment 16 for more information on behaviour we saw from Doppelganger during the reporting period).
Enforce our Community Standards
Work with fact-checkers
Strengthen our engagement with local experts and governments in the Central and Eastern Europe region

[Israel - Hamas War]

In the wake of the 07/10/2023 terrorist attacks in Israel and Israel’s response in Gaza, expert teams from across Meta took immediate crisis response measures, while protecting people’s ability to use our apps to shed light on important developments happening on the ground. As we did so, we were guided by core human rights principles, including respect for the right to life and security of the person, the protection of the dignity of victims, and the right to non-discrimination - as well as balancing those with the right to freedom of expression. We looked to the UN Guiding Principles on Business and Human Rights to prioritise and mitigate the most salient human rights risks: in this case, that people may use Meta platforms to further inflame an already violent conflict. We also looked to international humanitarian law (IHL) as an important source of reference for assessing online conduct. We have provided a public overview of our efforts related to the war in our Newsroom, as well as in our 2023 Annual Human Rights report. The following are some examples of the specific steps we have taken:

Taking Action on Violating Content:

We quickly established a dedicated crisis response staffed with experts, including fluent Hebrew and Arabic speakers, to closely monitor and respond to this rapidly evolving situation in real time. This allows us to remove content that violates our Community Standards faster, and serves as another line of defence against misinformation.
We continue to enforce our policies around Dangerous Organisations and Individuals, Violent and Graphic Content, Hate Speech, Violence and Incitement, Bullying and Harassment, and Coordinating Harm.

Safety and Security:

In addition to this, our teams detected and removed a cluster of Coordinated Inauthentic Behaviour (CIB) activity attributed to Hamas in 2021. These fake accounts attempted to re-establish their presence on our platforms.
In early 2025, we removed 17 accounts on Facebook, 22 FB Pages and 21 accounts on Instagram for violating our CIB policy. This network originated in Iran and targeted Azeri-speaking audiences in Azerbaijan and Turkey. Fake accounts – some of which were detected and disabled by our automated systems prior to our investigation – were used to post content, including in Groups, manage Pages, and to comment on the network’s own content – likely to make it appear more popular than it was. Many of these accounts posed as female journalists and pro-Palestine activists. The operation also used popular hashtags like #palestine, #gaza, #starbucks, #instagram in their posts, as part of its spammy tactics in an attempt to insert themselves in the existing public discourse.
We memorialise accounts when we receive a request from a friend or family member of someone who has passed away, to provide a space for people to pay their respects, share memories and support each other.

Reducing the Spread of Misinformation:

We’re working with third-party fact-checkers in the region to debunk false claims. Meta’s third-party fact-checking network includes coverage in both Arabic and Hebrew, through AFP, Reuters and Fatabyyano. When they rate something as false, we move this content lower in Feed so fewer people see it.
We recognise the importance of speed in moments like this, so we’ve made it easier for fact-checkers to find and rate content related to the war, using keyword detection to group related content in one place.
We’re also giving people more information to help them decide what to read, trust, and share, by adding warning labels on content rated false by third-party fact-checkers and applying labels to state-controlled media publishers.
We also have limits on message forwarding and we label messages that haven’t originated with the sender so people are aware that something is information from a third party.

User Controls:

We continue to provide tools to help people control their experience on our apps and protect themselves from content they don’t want to see. These include but aren’t limited to:

Hidden Words: This tool filters offensive terms and phrases from DM requests and comments.
Limits: When turned on, Limits automatically hide DM requests and comments on Instagram from people who don’t follow you, or who only recently followed you.
Comment controls: You can control who can comment on your posts on Facebook and Instagram and choose to turn off comments completely on a post by post basis.
Show More, Show Less: This gives people direct control over the content they see on Facebook.
Facebook Reduce: Through the Facebook Feed Preferences settings, people can increase the degree to which we demote some content so they see less of it in their Feed.
Sensitive Content Control: Instagram’s Sensitive Content Control allows people to choose how much sensitive content they see in places where we recommend content, such as Explore, Search, Reels and in-Feed recommendations.

Policies and Terms and Conditions

Outline any changes to your policies

War of Aggression by Russia on Ukraine

Policy
No further policy updates since our benchmark report.

Rationale
We continue to enforce our Community Standards and prioritise people’s safety and well-being through the application of these policies alongside Meta’s technologies, tools and processes. There are no substantial changes to report on for this period.

Israel - Hamas War

For the duration of the ongoing crisis, Meta has taken various actions to mitigate the possible content risks emerging from the crisis. This includes, inter alia, under the Dangerous Organisations and Individuals Policy, removes imagery depicting the moment an identifiable individual is abducted, unless such imagery is shared in the context of condemnation or a call to release, in which case we allow with a Mark as Disturbing (MAD) interstitial; and, remove Hamas-produced imagery for hostages in captivity in all contexts. Meta has some further discretion policies which may be applied when content is escalated to us.

Scrutiny of Ads Placements

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

War of aggression by Russia on Ukraine

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools and processes.

Measures taken to demonetise disinformation related to the crisis (Commitment 1 and Commitment 2)

As mentioned in our baseline report, our Advertising Standards prohibit ads that include content debunked by third-party fact-checkers and advertisers that repeatedly attempt to post content rated by fact-checkers may also incur restrictions to advertise across Meta technologies.For the monetisation of initially organic content, (1) per our Content Monetisation Policies, any content that's labelled as false by our third-party fact-checkers is ineligible for monetisation, and (2) any actor found in violation of our Community Standards, including our misinformation policies, may lose the right to monetise their content, per our Partner Monetisation Policies.

As mentioned in our baseline report, we prohibited ads or monetisation from Russian state-controlled media. Before Russian authorities blocked access to Facebook and Instagram, we paused ads targeting people in Russia, and advertisers in Russia are no longer able to create or run ads anywhere in the world.

Political Advertising

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

Depict a real person as saying or doing something they did not say or do; or
Depict a realistic-looking person that does not exist or a realistic-looking event that did not happen, or alter footage of a real event that happened; or
Depict a realistic event that allegedly occurred, but that is not a true image, video or audio recording of the event.

Meta will add information on the ad when an advertiser discloses in the advertising flow that the content is digitally created or altered. This information will also appear in the Ad Library. If it is determined that an advertiser did not disclose as required, Meta will reject the ad. Repeated failure to disclose may result in penalties against the advertiser.

The AI Disclosure policy helps inform people about digitally altered or created Ads. This way, people will be more aware about the authenticity of messaging, which will help combat Disinformation.

Integrity of Services

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

As mentioned in our baseline report, we have technical teams building scaled solutions to detect and prevent these behaviours, and are partnering with civil society organisations, researchers, and governments to strengthen our defences. We also improved our detection systems to more effectively identify and block fake accounts, which are the source of a lot of the inauthentic activity.

Since the invasion began, we shared what measures we’ve taken to help keep Ukrainians and Russians safe, our approach to misinformation, state-controlled media and ensuring reliable access to trusted information.As mentioned in our baseline report, our security teams took down three distinct networks in Russia targeting discourse on the war (announced here, here, and here) and have continued to monitor and enforce against Russian threat actors engaged in coordinated inauthentic behaviour (CIB). The Q4 2024 Adversarial Threat Report shared information on the continued low efficacy of the Doppelganger operation’s efforts on our apps, with most attempts to acquire fake accounts or run ads being quickly detected and blocked.

Relevant changes to working practices to respond to the demands of the crisis situation and/or additional human resources procured for the mitigation of the crisis (Commitment 14 -16)

As mentioned in the baseline report, throughout the war, we have mobilised our teams, technologies and resources to combat the spread of harmful content, especially disinformation and misinformation as well as adversarial threat activities such as influence operations and cyber-espionage.
We continue to work with a cross-functional team of experts from across the company, including native Ukrainian and Russian speakers, who are monitoring the platform around the clock, allowing us to respond to issues in real time.

Israel - Hamas War

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools and processes.

Removing a Coordinated Inauthentic Behaviour Network (Commitment 14, Commitment 16)
In early 2025, we removed 17 accounts on Facebook, 22 FB Pages and 21 accounts on Instagram for violating our policy against coordinated inauthentic behavior. This network originated in Iran and targeted Azeri-speaking audiences in Azerbaijan and Turkey. Fake accounts – some of which were detected and disabled by our automated systems prior to our investigation – were used to post content, including in Groups, manage Pages, and to comment on the network’s own content – likely to make it appear more popular than it was. Many of these accounts posed as female journalists and pro-Palestine activists. The operation also used popular hashtags like #palestine, #gaza, #starbucks, #instagram in their posts, as part of its spammy tactics in an attempt to insert themselves in the existing public discourse.

We removed this network before it was able to build authentic audiences on our apps.

Empowering Users

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

State controlled media: We continue to take the actions we outlined in our benchmark report. We have taken further action to limit the impact of state controlled media, described above.
Escalation channel: This channel continues to operate as outlined in our benchmark report.
Covert influence campaigns: We have continued to monitor for and remove recidivist attempts by coordinated inauthentic behaviour (CIB) networks that target discourse about the war in Ukraine. This covert activity is aggressive and persistent, constantly probing for weak spots across the internet, including setting up hundreds of new spoof news organisation domains.

Promotion of authoritative information, including via recommender systems and products and features such as banners and panels (Commitment 19)
As mentioned in our baseline report, we provided tools to help our community access crucial resources and take action to support people in need.

We continued supporting the Halo Trust and the State Emergency Service of Ukraine to spread authoritative factual information about the risks in contaminated areas, risks related to unexploded ordinances and life-saving information around shelters. Notably we sponsored the targeted ads campaigns of Halo Trust and improved the WhatsApp chat bot run by the State Emergency Service of Ukraine to ensure a safe and secure infoline.In addition, we provided an ad credits budget to 'Ty Yak?', a national mental health awareness campaign, to promote mental health resources for people affected by the war.

We continue to see funds raised on Facebook and Instagram for nonprofits in support of humanitarian efforts for Ukraine.

We continue to work through our Data for Good program, which empowers humanitarian organizations, researchers, UN agencies, and European policymakers to make more informed decisions on how to support the people of Ukraine.

Israel - Hamas War

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools, and processes.

Warning Screens on sensitive content, Sensitive Content Control and Facebook Reduce: (Commitment 17)
The 07/10/2023 attack by Hamas was designated as a Terrorist Attack under Meta’s Dangerous Organisation and Individuals policy. Consistent with that designation, we removed all content showing identifiable victims at the moment of the attack. Following that, people began sharing this type of footage in order to raise awareness and condemn the attacks. Meta’s goal is to allow people to express themselves while still removing harmful content. In turn, we began allowing people to post this type of footage within that context only, with the addition of a warning screen to inform users that it may be disturbing. If the user’s intent in sharing the content is unclear, we err on the side of safety and remove it.

However, there are additional protections in place to ensure people have choices when it comes to this content.

Instagram’s Sensitive Content Control allows people to choose how much sensitive content they see in places where we recommend content, such as Explore, Search, Reels and in-Feed recommendations. We try not to recommend sensitive content in these places by default, but people can also choose to see less, to further reduce the possibility of seeing this content from accounts they don’t follow.

Through the Facebook Feed Preferences settings, people can increase the degree to which we demote some content so they see less of it in their Feed. Or if preferred, they can turn many of these demotions off entirely. They can also choose to maintain Meta’s current demotions.

These actions ensure that we balance the protection of voice with removing harmful content. In this context, it has allowed for important discussion and condemnation of violence, while also empowering people to make choices in reaction to the content they see on Facebook and Instagram.

Hidden words Filter (Commitment 18, Commitment 19)
When turned on, Hidden Words filters offensive terms and phrases from DM requests and comments, so people never have to see them. People can customise this list, to make sure the terms they find offensive are hidden.

Hidden Words help people choose offensive terms and phrases to hide, so they are protected from seeing them.

Limits (Commitment 18, Commitment 19,)
When turned on, Limits automatically hide DM requests and comments on Instagram from people who don’t follow you, or who only recently followed you.

This tool gives people choice about DM and requests they receive, which may be important when engaging online around sensitive topics.

Comment Controls (Commitment 18, Commitment 19)
People can control who can comment on their posts on Facebook and Instagram and choose to turn off comments completely on a post by post basis.

This tool gives people control over engagement with what they post on Facebook and Instagram.

Show more Show less: (Commitment 18, Commitment 19)
Show More, Show Less gives people direct control over the content they see on Facebook. Selecting “Show more” will temporarily increase the amount of content that is like the post a user gave feedback on, while selecting “Show Less” means a user will temporarily see fewer posts like the one that feedback was given on.

This tool provides people with more direct control over what they see, which is important for protecting people's well-being during high profile crisis events.

Empowering the Research Community

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

War of Aggression by Russia on Ukraine

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools and processes.

Measures taken to support research into crisis related misinformation and disinformation (Commitment 17-25)
As mentioned in our baseline report, the Data for Good program shares privacy-protected data externally to help tackle social issues like disasters, pandemics, poverty and climate change. In support of the Ukraine humanitarian response, the program's maps have been utilized to provide valuable assistance.

As mentioned in our baseline report, we continued providing baseline population density maps (the high resolution settlement layer) of Ukraine and surrounding countries to humanitarian organisations for supply-chain planning and to aid demining efforts. These are the most accurate in the world with 30 metre resolution and demographic breakouts by combining updated census estimates with satellite imagery (i.e., no Facebook user data).

Our Social Connectedness Index has been used by leading researchers, including the European Commission - Joint Research Centre unit on Demography, Migration and Governance to quantify the rate at which Ukrainian refugees seek shelter in European regions with existing Ukrainian diaspora.

Israel - Hamas War

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools, and processes.

Content Library and API tools (Commitment 26)
As we previously reported, Meta has opened access to tools such as the Content Library and API tools to provide access to near real-time public content from Pages, Posts, Groups and Events on Facebook and public content on Instagram. Details about the content, such as the number of reactions, shares, comments and, for the first time, post view counts are also available. Researchers can search, explore and filter that content on both a graphical User Interface (UI) or through a programmatic API. Together, these tools provide the most comprehensive access to publicly-accessible content across Facebook and Instagram of any research tool built to date.

Individuals from qualified institutions, including journalists that are pursuing scientific or public interest research topics are able to apply for access to these tools through partners with deep expertise in secure data sharing for research, starting with the University of Michigan’s Inter-university Consortium for Political and Social Research. This is a first-of-its-kind partnership that will enable researchers to analyse data from the API in ICPSR’s Social Media Archives (SOMAR) Virtual Data Enclave.

Qualified individuals pursuing scientific or public interest research, including journalists can gain access to the tools if they meet all the requirements.

Empowering the Fact-Checking Community

Outline approaches pertinent to this chapter, highlighting similarities/commonalities and differences with regular enforcement.

War of Aggression by Russia on Ukraine

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools, and processes.

Cooperation with independent fact-checkers in the crisis context, including coverage in the EU (Commitment 30-33)
As mentioned in our baseline report, for misinformation that does not violate our Community Standards, but undermines the authenticity and integrity of our platform, we work with our network of independent third-party fact-checking partners.

The details of the network are outlined under the Empowering Fact-Checkers chapter above.

As mentioned in our baseline report, our cooperation with fact-checkers is as outlined in the Fact-Checkers’ Empowerment chapter above.

In Europe, we partner with 46 fact-checking organisations, covering 36 languages. This includes 29 partners covering 26 countries and 23 different languages in the EU.

Israel - Hamas War

As noted in our baseline report, our policies are based on years of experience and expertise in safety combined with external input from experts around the world. We are continuously working to protect the integrity of our platforms and adjusting our policies, tools, and processes.

Working with fact checker in the region and deploying keyword detection (Commitment 30)
Meta is working with third-party fact-checkers in the region to debunk false claims. Meta’s third-party fact-checking network includes coverage in both Arabic and Hebrew, through AFP, Reuters and Fatabyyano. We recognise the importance of speed in moments like this, so we’ve made it easier for fact-checkers to find and rate content related to the war, using keyword detection to group related content in one place.

When they rate something as false, we move this content lower in Feed so fewer people see it.

Content Warning Labels Commitment 31)
Meta is adding warning labels on content rated false by third-party fact-checkers and applying labels to state-controlled media publishers. We also have limits on message forwarding and label messages that haven’t originated with the sender so people are aware that something is information from a third party.

Meta is supporting people in the region by giving them more information to decide what to read, trust and share by adding warning labels onto relevant content.