Changes from report parliamentary committee draft to plenary report

– having regard to the Influencer Handbook, produced by Finland’s Mediapooli in 2020,

– having regard to the report by the US Department of State of August 2020 entitled ‘Pillars of Russia’s Disinformation and Propaganda Ecosystem’,

Changed– having regard to the joint cybersecurity advisory by the US Cybersecurity and Infrastructure Security Agency, the US National Security Agency, the FBIUS Federal Bureau of Investigation (FBI) and international partners of August 2025 entitled ‘Countering Chinese State-Sponsoredstate-sponsored Actorsactors Compromisecompromise of Networksnetworks Worldwideworldwide to Feed Globalglobal Espionageespionage Systems’,systems’,

– having regard to the report by Reporters Without Borders of 25 September 2025 entitled ‘The Propaganda Monitor: The Russian Edition’,

Changed– having regard to the article published by the Centre for Media Pluralism and Media Freedom at the having regard to the article published by the Centre for Media Pluralism and Media Freedom at the European University Institute on 1 November 2025 entitled ‘Influencers as news creators: implications for media regulation’,

– having regard to the article published by the AlgoSoc Consortium on 28 October 2025 entitled ‘1 in 10 Dutch citizens are likely to ask AI for election advice. This is why they shouldn’t’,

Changed– having regard to the report by What to Fix of June 2025 entitled ‘Bankrolling Sanctionedsanctioned Entities:entities: How Meta Platforms Ireland Ltd. Maymay Havehave Violatedviolated EU Sanctionssanctions and Channeledchanneled Moneymoney To RT, Sputnik and Otherother EU-SanctionedEU-sanctioned Entitiesentities via Facebook’s Revenuerevenue Redistributionredistribution Programs’,programs’,

Changed– having regard to the special report by NewsGuard of 6 March 2025 entitled ‘A Well-fundedwell-funded Moscow-based Globalglobal “News”“news” Networknetwork has Infectedinfected Western Artificialartificial Intelligenceintelligence with Russian Propaganda’,propaganda’,

Changed– having regard to the report by Media Freedom Rapid Response entitled ‘Mapping Media Freedom – Monitoring Report 2024’,2025’, published in February 2025,2026,

– having regard to the report by the World Economic Forum of January 2025 entitled ‘Global Cybersecurity Outlook 2025’,

Changed– having regard to the investigative report published by VSquare, Delfi Estonia and partner media organisations on 26 February 2024 entitled ‘Kremlin Leaks:leaks: Secret Filesfiles Revealreveal Howhow Putin Pre-Riggedpre-rigged his Reelection’,reelection’,

▸4 unchanged paragraphs

– having regard to the Ethical Code of Conduct for Social Media Influencers and Content Creators, published by the Aspen Institute Germany in 2024,

– having regard to the article published by Debunk.org on 4 May 2023 entitled ‘Kremlin spent 1.9 billion USD on propaganda last year, the budget exceeded by a quarter’,

– having regard to the statement by the Russian Federation’s Ministry of Foreign Affairs of 28 December 2024 on measures in response to the EU’s 15th sanctions package against Russia, in which it announced the expansion of the list of EU officials and citizens prohibited from entering the country,

– having regard to the article by the Russian News Agency TASS of 28 December 2024 entitled ‘Russia substantially expands blacklist of EU officials in response to sanctions – MFA’,

Added– having regard to the Commission press release of 18 December 2023 on its opening of formal proceedings under the Digital Services Act against X to assess possible breaches of its obligations, in particular as regards the mitigation of systemic risks linked to the design and functioning of its recommender systems, as well as obligations related to risk assessment, content moderation, advertising transparency and access to data for researchers,

Added– having regard to a range of academic and investigative studies on social media recommender systems in the context of the German federal elections, which indicate algorithmic bias and the amplification of polarising or partisan content, thereby shaping users’ political information environment,

Added– having regard to the Commission press release of 30 April 2024 on its opening of formal proceedings under the Digital Services Act against Meta platforms, in particular as regards its handling of political content on its online platforms,

– having regard to the statement by TikTok of December 2024 on continuing to protect the integrity of TikTok during Romanian elections,

Changed– having regard to the statement issued by the European Solar Manufacturing Council on 30 April 2025 entitled ‘Restrict Remoteremote Accessaccess of PV Invertersinverters from High-Riskhigh-risk Vendors’,vendors’, in which it warned of the risks to Europe’s energy sovereignty duebecause toof unregulated and remote-control capabilities of PV inverters from high-risk, non-European manufacturers,

Removed– having regard to Rule 55 of its Rules of Procedure,

Added– having regard to the White Paper for European Defence – Readiness 2030 presented on 19 March 2025 and Council Regulation (EU) 2025/1106 of 27 May 2025 establishing the Security Action for Europe (SAFE) through the Reinforcement of the European Defence Industry Instrument,

Changed– having regard to the reportCouncil presidency conclusions of the28 SpecialNovember Committee2025 on access to reliable news as part of the European Democracy Shield (A10-0000/2025),Shield,

Change 1

RemovedA. whereas on 12 November 2025, the Commission and the High Representative of the Union for Foreign Affairs and Security Policy presented a joint communication on the establishment of the European Democracy Shield, setting out a series of measures to empower, protect and promote strong and resilient democracies across the EU;

Added– having regard to Rules 55 and 213 of its Rules of Procedure,

RemovedB. whereas the European Centre for Democratic Resilience constitutes the centrepiece of the European Democracy Shield; whereas both Parliament’s Special Committees on Foreign Interference in all Democratic Processes in the European Union, including Disinformation (INGE and INGE 2), had already called for the establishment of an EU structure to counter foreign information manipulation and interference (FIMI);

Added– having regard to the report of the Special Committee on the European Democracy Shield (A10-0199/2026),

RemovedC. whereas recent EEAS reports on FIMI threats demonstrate an increasingly granular understanding of malicious actor methodologies, technical infrastructure and intended narratives, yet this situational awareness has not been translated into sustained operational mechanisms that enable the EU to take timely, coordinated countermeasures; whereas the EU’s institutional nexus is a principal focus of hostile information manipulation campaigns; whereas the EU is suffering from a fragmented approach to countering FIMI and disinformation, with significant variations in Member States’ capacities and no overarching strategic framework linking operational response to threat intelligence;

AddedA. whereas on 12 November 2025, the Commission and the High Representative of the Union for Foreign Affairs and Security Policy presented a joint communication on the establishment of the European Democracy Shield, setting out a series of measures to empower, protect and promote strong and resilient democracies across the EU; whereas this initiative provides a unique opportunity to build an effective, whole-of-society approach to the protection of European democracy, freedom and prosperity; whereas democratic resilience and European security are increasingly inseparable;

RemovedD. whereas the ECNE, the Rapid Alert System (RAS) and the European Digital Media Observatory (EDMO) are valuable components of the EU’s overall resilience to FIMI and disinformation, yet their effectiveness is substantially limited by the absence of a dedicated EU operational structure with the authority to coordinate resilience-building activities and rapid response and escalation mechanisms connecting national operational capacities to EU-level coordination;

AddedB. whereas poverty, socio-economic deprivation and marginalisation can leave certain segments of society particularly exposed and vulnerable to external interference and hostile information operations;

RemovedE. whereas the threats posed by FIMI and disinformation extend beyond individual Member States, to an attack on the very essence of the European project; whereas FIMI operations systematically target the core democratic values of Article 2 TEU and the principle of open, transparent decision-making processes enshrined in Article 1 TEU, in a manner that adversely affects the interests of the EU;

AddedC. whereas the threat posed by Russia to Europe’s security and democratic integrity remains at unprecedented levels; whereas other actors, including China and Iran, also pose significant challenges; whereas it is necessary to adopt a principle of information deterrence and, where appropriate, apply restrictive measures against actors responsible for foreign information manipulation and interference (FIMI);

RemovedF. whereas individual Member States that have invested in dedicated operational structures, with clear statutes and mandates, as well as sufficient funding and staffing, demonstrate what is achievable, most notably in the cases of France’s VIGINUM and Sweden’s Psychological Defence Agency;

AddedD. whereas the joint communication on the European Democracy Shield contains several positive recommendations and proposals, such as the establishment of an European Centre for Democratic Resilience, the extension of its scope to accession countries and the EU neighbourhood and the full enforcement of key EU legislation; whereas both of Parliament’s Special Committees on Foreign Interference in all Democratic Processes in the European Union, including Disinformation (INGE and INGE 2) had already called for the establishment of an EU structure similar to the proposed European Centre for Democratic Resilience to counter FIMI and disinformation; whereas such a structure should not duplicate existing tools, capabilities and competences, but instead consolidate them into a single entity with a clear vision and mandate, real powers and financing;

RemovedG. whereas the rapid advancement of artificial intelligence (AI) and deepfake technologies is outpacing the adaptive capacity of fragmented national institutions; whereas without coordinated EU responses and clear operational capabilities, the technological advantage will increasingly favour attackers over defenders;

AddedE. whereas recent EEAS reports on FIMI attacks demonstrate an increasingly granular understanding of malicious actor methodologies, technical infrastructure and intended narratives, yet this situational awareness has not been translated into sustained operational mechanisms that enable the Member States and the EU to anticipate, prevent, detect and take timely, coordinated countermeasures; whereas national and European elections are the principal focus of hostile information manipulation campaigns and other forms of malign attacks; whereas these campaigns disparage particular candidates or political parties, attempt to discredit the electoral process itself, instrumentalise divisive issues and attempt to discredit traditional media; whereas EU and Member States’ institutions are also intensively attacked; whereas the EU is suffering from a fragmented approach to countering FIMI and disinformation, with significant variations in Member States’ capacities and no overarching strategic framework or functioning transnational or European cooperation mechanisms linking operational response to threat intelligence;

RemovedH. whereas the concurrence of the early stages of the implementation of the Digital Services Act (DSA), the Digital Markets Act (DMA), the Artificial Intelligence Act (AI Act) and the European Media Freedom Act (EMFA) establishes a window of opportunity to build comprehensive FIMI defences;

AddedF. whereas ECNE, the Rapid Alert System and the European Digital Media Observatory (EDMO) are valuable components of the EU’s overall resilience to FIMI and disinformation, yet their effectiveness is substantially limited by the absence of a dedicated EU operational structure with the authority to coordinate resilience-building activities and rapid response and escalation mechanisms that connect national operational capacities to EU-level coordination;

RemovedI. whereas democratic societies in the EU are increasingly being targeted by hybrid threats, disinformation and FIMI, with a particular intensity in the digital sphere; whereas the online space enables the proliferation of new manipulation techniques, including: (a) the inauthentic use of social media through automated software programmes, fake social media accounts and the use of troll factories, bot-driven amplification and engagement, (b) the use of astroturfing and flooding techniques to influence online public debate, (c) the personalisation, tracking and micro-targeting of individuals, (d) websites designed to mimic official sources, (e) the artificial amplification of divisive content, (f) the use of synthetic content, such as deepfakes and other AI-generated content, and (g) recommender systems purposefully designed to drive engagement through polarisation;

AddedG. whereas the threats posed by FIMI and disinformation extend beyond individual Member States, to an attack on the very essence of the European project; whereas FIMI operations systematically target the core democratic values of Article 2 TEU and the principle of open, transparent decision-making processes enshrined in Article 1 TEU, in a manner that adversely affects the interests of the EU and democratic processes in its Member States; whereas, ultimately, European citizens are the primary victims of these practices, as their access to reliable information and their capacity to form independent opinions and to participate meaningfully in democratic life are directly weakened, while public security and societal stability are also put at risk;

RemovedJ. whereas online platforms, when deciding on whether or not to promote content in a user’s feed, use their insights into the preferences and vulnerabilities of each individual user, making them more influential than traditional publishers;

AddedH. whereas some individual Member States that have invested in dedicated operational structures, with clear statutes and mandates, as well as sufficient funding and staffing, demonstrate what is achievable, most notably in the cases of France’s VIGINUM and Sweden’s Psychological Defence Agency, which have adopted a fact-based, technical and non-partisan approach; whereas these institutions contribute valuable insights on request or on their own initiative but are not appropriately interlinked;

RemovedK. whereas the EU has recently adopted a set of comprehensive legislative acts to establish a safe and transparent online space, including the DSA, the regulation on the transparency and targeting of political advertising, and the AI Act; whereas these laws represent an important step forward in holding digital platforms accountable and protecting democratic processes, increasing transparency and safeguarding freedom of expression; whereas, at the same time, their enforcement remains a challenge;

AddedI. whereas the rapid advancement of artificial intelligence (AI) and deepfake technologies is outpacing the adaptive capacity of fragmented national institutions and making old-school solutions, such as transparency or best efforts obligations obsolete; whereas without coordinated EU responses and clear operational and research capabilities, the technological advantage will increasingly favour attackers over defenders, thus threatening the core basis of our democracies; whereas some AI companies have scaled back, instead of increasing, their safety measures to counter the spread and generation of disinformation through their services; whereas AI facilitates disinformation campaigns but it also provides tools to improve our capacity to detect, analyse and counter them;

RemovedL. whereas the use of AI presents new opportunities for strengthening election management and oversight, such as through detecting unusual online activity and coordinated manipulative behaviour; whereas the use of AI, including generative models and deepfakes, also presents significant challenges for information integrity and electoral processes; whereas credible reports have exposed the operations of a Moscow-based disinformation network known as ‘Pravda’, which has deliberately targeted and infiltrated large language models (LLMs) with pro-Kremlin narratives designed to mimic neutral and fact-based outputs; whereas such tactics represent a dangerous evolution of FIMI, as they exploit the opacity and scalability of AI systems to subtly disseminate false or misleading geopolitical messaging under the guise of authoritative language; whereas this form of algorithmic manipulation threatens to erode public trust in AI technologies, distort democratic discourse, and exacerbate disinformation risks across the EU and globally; whereas this underlines the need for strengthened AI literacy in education, work and society at large;

AddedJ. whereas the principle that industries should contribute financially to addressing externalities related to their activities is well-established in Union law and practice; whereas under the Extended Producer Responsibility framework, certain producers are required to finance the management of the waste generated by their products, in direct application of the polluter-pays principle; whereas under the Single Resolution Mechanism, all banks operating in the banking union make annual contributions to the Single Resolution Fund so that the systemic risks related to their activities do not fall on taxpayers; whereas pharmaceutical companies are required to fund post-authorisation safety studies, on the basis that those who profit from placing a product on the market must contribute to monitoring the harms it may cause; whereas this logic should apply to large information society service providers who benefit from the information ecosystem while simultaneously creating systemic risks to democratic integrity;

RemovedM. whereas online influencers can play an important role in strengthening the digital resilience of our societies, when raising awareness about media literacy, democratic procedures or contributing to a healthy political debate;

AddedK. whereas the EU has, on several occasions, successfully navigated legally contested and emerging policy areas through structured preparatory instruments; whereas notable examples include the 1988 Green Paper on Copyright and the Challenge of Technology, the High-Level Expert Group on Artificial Intelligence, and the pre-legislative consultation process preceding the General Data Protection Regulation, among others; whereas FIMI and disinformation present a comparable challenge related to technological, international relations and business models evolving together with societal and legal phenomena while also competing with constitutional considerations; whereas a tailored multidisciplinary preparatory process could be critical for a robust, court-proof legislative framework related to FIMI and disinformation in full respect of Member States’ competences and the powers of the EU institutions;

RemovedN. whereas the autumn 2024 presidential elections in Romania exposed significant vulnerabilities in the online information environment, with credible reports indicating the widespread use of fake accounts, bots and algorithmically amplified content to promote specific political narratives; whereas TikTok, one of the most widely used platforms among young voters, claimed to have blocked over 116 000 spam accounts from being created and removed 59 000 fake accounts in a post-election clean-up; whereas the use of influencers, including nano- and micro-influencers, by foreign actors further highlights the fact that low-cost, low-visibility actors can be weaponised to evade detection and influence public opinion, in breach of democratic norms;

AddedL. whereas democratic societies in the EU are increasingly being targeted by hybrid threats, disinformation and FIMI, with a particular intensity in the digital sphere; whereas the online space enables the proliferation of new manipulation techniques, including: (a) the inauthentic use of social media through automated software programmes, fake social media accounts and the use of troll factories, bot-driven amplification and engagement, (b) the use of astroturfing and flooding techniques to influence online public debate, (c) the personalisation, tracking and micro-targeting of individuals, (d) websites designed to mimic official sources, (e) the artificial amplification and monetisation of divisive content, (f) the use of synthetic content, such as deepfakes and other AI-generated content, (g) recommender systems designed to drive engagement through polarisation and (h) LLM (large language model) grooming and other data poisoning techniques maliciously feeding AI systems with biased, deceptive or false narratives to influence its training; whereas tools to counter these techniques and negative phenomena have been developed but remain widely insufficient;

RemovedO. whereas Telegram, a messaging platform with a growing user base in the EU, has increasingly been flagged as a conduit for disinformation, foreign interference and illicit activities, including the spread of extremist content and untraceable political propaganda; whereas its origins in Russia and lack of meaningful content moderation have raised persistent concerns regarding its role in undermining democratic discourse; whereas Telegram has become a primary channel for recruiting ‘disposable agents’ and conducting hybrid operations; whereas malicious state actors have been found to exploit the technological skills, financial needs and naivety of Telegram users, especially young people, to conduct espionage and sabotage operations at minimal cost, while maintaining plausible deniability; whereas reports suggest that Telegram meets the user threshold for it to be classified as a Very Large Online Platform (VLOP) under the DSA, making it subject to stricter regulatory obligations and risk-mitigation requirements;

AddedM. whereas online platforms, through insights into the preferences and vulnerabilities of each individual user, wield greater influence over public discourse than traditional publishers; whereas evidence suggests that platforms’ algorithms can be manipulated by their owners and shareholders to prioritise certain political content, thereby distorting the information environment and undermining the possibility of users receiving accurate and pluralistic information; whereas very large online platforms (VLOPs) and very large search engines (VLOSEs) have become de facto gatekeepers of the public space, thereby strongly influencing what information citizens encounter, how it is prioritised, and under what conditions it circulates;

RemovedP. whereas the growing influence of non-European technology companies – particularly those based in the United States and China – has highlighted the EU’s strategic dependencies in critical areas such as data infrastructure, cloud services and social media platforms; whereas fostering home-grown innovation, supporting European start-ups and investing in sovereign digital infrastructure are essential steps towards building a competitive, secure and resilient EU digital ecosystem; whereas aligning digital innovation with democratic values and fundamental rights offers the EU an opportunity to become a global leader in responsible and human-centred technology;

AddedN. whereas the EU and its Member States have recently adopted a set of comprehensive legislative acts to establish a safe and transparent online space, including the Digital Services Act (DSA), the Digital Markets Act (DMA), the Transparency and Targeting of Political Advertising Regulation (TTPA), the European Media Freedom Act (EMFA) and the Artificial Intelligence Act (AI Act); whereas these acts mark important progress in holding online platforms accountable and protecting democratic processes, increasing transparency, safeguarding freedom of expression, and thus contributing to comprehensive FIMI defences, but their effectiveness depends on full, timely and consistent enforcement, which remains challenging and slow, creating vulnerabilities; whereas structured support for the Member States in this regard is therefore particularly important; whereas the EU and its Member States should also continue to develop non-legislative measures, common approaches and methodologies in order to disrupt FIMI and defend democracy; whereas Parliament plays a strong role in scrutinising and reinforcing the EU’s digital legislation, and these efforts should be complemented by regular reviews of Parliament’s working methods to ensure effectiveness;

RemovedQ. whereas research shows a concerning trend in the revenue redistribution programmes of online platforms, revealing that media outlets that are subject to sanctions, such as Sputnik and Russia Today, remained listed as ‘partner-publishers’ on Facebook months after the launch of EU sanctions against Russia – raising questions about whether they have continued to benefit financially; whereas such opaque monetisation mechanisms enable foreign state‑affiliated actors to profit from sharing disinformation, and continue to destabilise EU information spaces despite regulatory measures;

AddedO. whereas evidence demonstrates that the engagement-based recommendation systems of VLOPs, including X, TikTok and Instagram, are structurally driven by attention-based economics to amplify divisive, polarising or misleading content, thereby posing risks to civic discourse and electoral integrity, as evidenced by the amplification of extremist political content during elections across the EU; whereas these systems have undermined the visibility of editorially independent and pluralistic media, and facilitate the rapid spread of AI-generated deepfakes and synthetic or manipulated content, exacerbating the challenge for users, in particular vulnerable groups, of distinguishing between true and false information; whereas the AI Act introduced rules on mandatory labelling of AI-generated content; whereas Articles 34 and 35 DSA require VLOPs to assess and mitigate systemic risks; whereas the DSA also introduced significant new rules to increase algorithmic transparency and accountability, including the option for users to choose non-personalised feeds, thereby reducing reliance on opaque algorithmic curation; whereas challenges within the implementation and user experience remain; whereas effective enforcement of the DSA and the AI Act, supported by up-to-date guidelines and independent oversight, is essential to counter manipulation and safeguard democratic processes;

RemovedR. whereas in a survey held in the run-up to the 2025 parliamentary elections in the Netherlands, 1 in 10 respondents declared that they would be likely to ask AI for voting advice, while another 13 % declared they would maybe use it; whereas in the Netherlands, the combined ratio for the youngest voters amounts to more than one third of the voters; whereas this trend is consistent with findings across the Member States and globally; whereas a recent study by the Dutch data protection authority demonstrates that voting recommendations generated by AI chatbots often present a highly distorted and polarised view of the political landscape;

AddedP. whereas targeted sanctions, threats of tariffs, visa bans, and personal sanctions have been directed by the US administration at EU actors involved in shaping and enforcing the DSA and other digital legislation, undermining the integrity of EU decision-making and the rule of law; whereas the EU, as an open and rules-based single market, must uphold its sovereign right to legislate in line with democratic values, international commitments and the protection of fundamental rights, including freedom of expression, which is a core value shared with the United States and like-minded partners across the democratic world; whereas unfounded allegations of censorship undermine these shared values and divert attention from actual serious free speech violations in countries such as Russia, China or Iran; whereas the enforcement of EU digital legislation must resist delays or limitations resulting from geopolitical considerations, notably in cases involving lobbying by big tech platforms; whereas external pressures, including geoeconomic coercion, strategic acquisitions and support for extremist movements by non-EU public or corporate actors, further threaten the EU’s decision-making sovereignty and democratic resilience; whereas the enforcement of the EU’s digital rules must remain guided solely by Union law and the protection of democratic interests, free from the undue influence of political, trade or external pressures; whereas unity among the EU Member States and their political alignment in this regard are key;

RemovedS. whereas the fundamental right to freedom of expression and information, as enshrined in Article 11 of the Charter of Fundamental Rights of the European Union and Article 10 of the European Convention on Human Rights, is a cornerstone of democracy; whereas the EU’s commitment to freedom of expression represents a coherent and principled counter-FIMI strategy, in clear contrast with the restrictions that authoritarian actors systematically impose on their populations; whereas the freedom and pluralism of the media, and the existence of a vibrant civic space, empower societies to detect, expose and reject manipulative narratives through democratic deliberation;