EESC info

New research from the Balkan Free Media Initiative (BFMI) raises alarm about Europe’s preparedness to counter platform manipulation and election interference. In its latest report Tackling TikTokcracy: A blueprint for fighting algorithmic manipulation in Europe,  BFMI shows how TikTok and other platforms have been heavily exploited across recent elections in the Balkans, including through large-scale networks of fake accounts and cross-platform amplification. The report documents tactics such as hashtag hijacking and blended influencer-bot networks, revealing shared structural weaknesses that leave democratic processes vulnerable. For EESC Info, BFMI outlines concrete recommendations aimed at strengthening democracy and security ahead of future elections.

By the Balkan Free Media Initiative (BFMI)

The latest report by BFMI, developed in partnership with analytics firm Sensika, uncovered powerful networks of digital interference in Romania, Bulgaria and Kosovo which use sophisticated hybrid tactics to mimic genuine online engagement and artificially boost political messaging. These strategies are not confined to the Balkans, the report warns, and are spreading across Europe faster than institutions, platforms and citizens are currently responding.

The authors call for a rapid reimagining of Europe’s democratic defence architecture, one that makes platforms more transparent, proactively detects fake online activity across borders and builds citizens’ resilience to online influence. Without taking action, Europe risks becoming a ‘TikTokcracy’, in which algorithms – not citizens – decide its future.

Algorithmic influence exploits cracks in media systems

Across the Balkans, BFMI researchers found that networks of automated accounts, paid influencers and misinformed supporters were taking advantage of algorithmic incentives and regulatory gaps. Through sophisticated strategies like high-volume posting, hashtag engineering and the blending of political and entertainment content, those involved in these networks were, both knowingly and unknowingly, amplifying disinformation and manipulating public opinion.

The annulment of Romania’s presidential election in 2024 first revealed the scale of this threat, when such networks directly undermined electoral processes. Romanian intelligence exposed a large-scale operation, which coordinated over 25,000 automated TikTok accounts and a network of micro-influencers to artificially boost one candidate’s content to users and drive them to the polls in his favor. Much of this coordination took  place via Telegram, where locally-resonant narratives and strategic hashtags were distributed en masse. 

Similar dynamics were unearthed in Bulgaria, where politically-charged content disseminated from bogus websites and monetised through non-transparent advertising contributed to the country’s four years of perpetual election cycles.  The problem is perhaps even more worrying here, as the report unearthed a cross-platform amplification model that is economically embedded in Bulgaria’s captured media landscape, constantly adapting and ready to be activated at any time. 

In Kosovo, such tactics helped to create an especially tense campaign environment in 2025 with the potential to inflame pre-existing ethnic tensions at a particularly sensitive time. Once again, the shared strategies included the 'Fire Hose' tactic of mass posting and commenting, synchronisation of engagement, targeted hashtag usage, and the blending of unlabeled political advertisement with entertainment or sport content.

However, historic anti-corruption protests in Bulgaria via TikTok at the end of 2025 have illustrated that these platforms can be a double-edged sword for democracy, capable of both energising civic participation and undermining political stability.

A core finding of BFMI’s report is that the manipulation of platform algorithms thrives where and when media ecosystems are fragmented, non-transparent and captured by political or business interests. While these vulnerabilities are indeed prevalent in the Balkans, under-regulated platforms, poor transparency standards, fragile media infrastructure and limited cross-border cooperation are shared weaknesses across Europe. Without a concerted response spearheaded in Brussels, all Member States remain at risk of ‘TikTokcracy’.

Strengthening European defences against ‘TikTokcracy’

The report moves beyond diagnosis to outline a clear and immediate European policy framework which not only involves policymaking bodies but also includes efforts to mobilise society as a whole, from platforms to national institutions to European citizens. This includes: 

• creating forensic and monitoring tools under the European Democracy Shield that feed into an EU-wide early-warning and rapid-response system;

• aggressively enforcing existing legislation such as the Digital Services Act (DSA) and the European Media Freedom Act (EMFA), and creating additional binding platform guidelines for political content, advertising transparency and election protection;

• new financing, training and tech support for credible, independent media to assist in the fight against disinformation;

• digital literacy initiatives for young voters and public awareness campaigns to protect citizens from threats and rebuild societal trust.

Together, these measures would significantly strengthen Europe’s capacity to safeguard free and fair democratic debate and ensure that political agency remains with citizens rather than with platforms.

As BFMI’s findings make clear, algorithmic manipulation is evolving faster than Europe’s current defences, requiring more than incremental adjustments or regulatory enforcement, including a harmonised response that matches the speed and scale of these threats. Europe must either modernise its democratic defences for the digital age or risk allowing algorithmic visibility and manufactured popularity to erode public trust, distort political choice and weaken democratic societies.

The Balkan Free Media Initiative (BFMI) is a Brussels-based organisation that promotes media freedom and safeguards journalists’ rights in the Balkans. BFMI focuses on promoting transparency, accountability and ethical journalism, while countering disinformation, hybrid threats and other obstacles that undermine Euro-Atlantic values. By supporting collaboration among media professionals and civil society, BFMI helps reinforce democratic resilience. Thanks to its comprehensive approach, the initiative plays a crucial role in empowering independent media, ensuring that diverse voices are heard across the Balkans.

With more than 1.6 billion users globally, including over 200 million across Europe, TikTok has become a major space for political expression and information-sharing and a primary source of news for a significant share of young people. As a result, it has come under growing scrutiny from regulators and civil society. The European Commission opened formal proceedings in 2024 to examine whether TikTok is adequately assessing and mitigating systemic risks related to election integrity and civic discourse. We asked Francesca Scapolo, TikTok's Election Integrity Expert for Public Policy in Europe, how TikTok understands its responsibility for these risks in practice, how it cooperates with authorities, and what safeguards it has in place to protect democratic processes.

Given TikTok’s scale and its growing role as a source of political information for millions of users across the EU, how does the company approach accountability for systemic risks on the platform more broadly, such as the spread of disinformation, coordinated behaviour, or fake and inauthentic accounts? How do these efforts translate into cooperation with national authorities and EU institutions, particularly during sensitive moments like elections?

TikTok is a discovery platform where more than 200 million Europeans come to connect, share their passions, and find inspiration. We recognise that with scale comes responsibility, and we work continually to protect our platform and maintain a civil place for people to express themselves and build community, including during elections. We’ve invested significantly in systems, specialised teams and partnerships to address systemic risks such as harmful misinformation, fake and inauthentic accounts, coordinated inauthentic behaviour, and deceptive behaviours.

Across the EU, our work includes proactive enforcement of our Community Guidelines, investment in features, tools and resources to empower our community, including media literacy initiatives, and partnerships with external experts. In fact, through TikTok's global fact-checking programme, we work closely with more than 20 IFCN-accredited fact-checking organisations, including AFP in France, DPA in Germany and Newtral in Spain.

Our technical and enforcement work is complemented by ongoing cooperation with national authorities and EU regulators. Under the Digital Services Act (DSA) and the Code of Conduct on Disinformation, we engage with Digital Service Coordinators and the European Commission. We also provide regular updates on our content moderation efforts through our transparency reports.

During high‑stakes periods, such as elections, we also collaborate with national authorities and electoral commissions, and participate in the Code’s rapid response system, which enables swift, coordinated information-sharing between civil society organisations, fact‑checkers and platforms to address urgent or emerging threats, a critical capability during elections.

Taken together, these efforts demonstrate how we blend proactive risk mitigation, user‑empowering tools, and regulatory cooperation to help safeguard democratic discourse across the EU, especially during sensitive electoral moments.

From your perspective, are the measures TikTok currently has in place sufficient to address systemic risks to democratic processes during elections, particularly those linked to recommendation algorithms, visibility dynamics, and coordinated campaigns? Or do you see a need for stronger or more proactive safeguards?

During elections, we work continually to protect our platform and maintain a civil place for people to express themselves and build community. Thousands of trust and safety and security professionals have safeguarded TikTok through over 200 elections around the world over the last five years. Our comprehensive strategy is based on three key pillars:

• Protecting election integrity: Removing harmful misinformation, disrupting attempts to influence our community, including covert influence operations, collaborating with fact-checkers to assess content accuracy, and labelling unverifiable claims.

• Empowering users: Providing access to reliable information through Election Centres, enabling users to separate fact from fiction.

• Collaborating with experts: Partnering with electoral commissions and fact-checking organisations to counter emerging threats.

Through these efforts, in 2025, we disrupted more than 75 covert influence networks, and removed tens of thousands of accounts for violating our covert influence policies. We stay accountable to our community with regular updates on how we protect election integrity and frequent reports on the covert influence operations we have disrupted.

Looking ahead, we remain committed to strengthening these pillars and to evolving our safeguards as risks change.

Project Clover has been presented as a key pillar of TikTok’s European data-governance strategy, including a long-term investment of around EUR 12 billion, yet it remains relatively unknown to the public. How does this initiative concretely change how TikTok handles European user data, and what relevance does it have for election integrity and democratic safeguards in the EU?

Project Clover is one of the most advanced and comprehensive data protection programmes to be found anywhere. Its core tenets include storing European user data in a dedicated European enclave by default and putting additional safeguards and restrictions around that data, building on our existing controls on who can access data.

We've also engaged a respected European cybersecurity firm, NCC Group, to independently monitor and verify these safeguards. NCC Group's oversight provides third-party accountability over our work to protect European user data. We've also deployed tools to further protect European user privacy called 'privacy enhancing technologies'.

These measures go further than regulatory requirements, while being aligned with principles in the GDPR, and our general efforts towards safeguarding our platform and users through robust processes, policies, and procedures.

Francesca Scapolo oversees TikTok’s Europe-wide election integrity public policy efforts, coordinating among product, trust and safety, and policy teams. Collaborating with internal and external stakeholders, she implements regional public policy strategies that reinforce civic trust and safeguard electoral integrity. Before joining TikTok, she worked at the Meta Oversight Board and the European Commission. 

In an online world where generative AI can fabricate a headline, an image and a source in seconds, 'breaking news' may soon give way to 'fact-checked news'. At a time when lies travel faster than facts, fact-checking is quickly becoming one of journalism's most powerful tools. The European Digital Media Observatory (EDMO) tracks Europe's most persistent false narratives through its monthly disinformation briefs. We spoke with EDMO coordinator Tommaso Canetta about how fact-checking is evolving — and what it takes to push back against disinformation in the age of AI.

Could you tell us a little bit more about the monthly briefs of the EDMO fact-checking network? How do you collect information and decide what to include in the briefs? Who are your fact-checkers?

Every month, we send a questionnaire to the fact-checking organisations that are members of the EDMO fact-checking network (55 organisations covering all EU Member States plus Norway). The questionnaire includes both quantitative and qualitative questions about the disinformation detected during the previous month. We then analyse all the responses and include the most relevant information emerging from this analysis in the briefs.

Your October brief stated that AI-generated disinformation hit a new record amid the crumbling of information integrity. What is an 'AI slop' and how is it used to produce fake news or political discreditation? Can you give us some recent and blatant examples?

'AI slop' can be defined as low- to mid-quality content created using AI tools. The deluge of AI-generated content circulating on social media platforms during crises, before, during or after elections, and more generally around sensitive topics, can significantly distort public perception.

Recent examples include the many false videos and images allegedly showing Venezuelans celebrating in the streets following the abduction of Maduro by the United States. Another example is the circulation, in November, of AI-generated videos depicting Ukrainian soldiers surrendering. In the political sphere more broadly, deepfakes of politicians saying things they never said are increasingly being created and disseminated to discredit them (for example, this one in Hungary).

Are there some recurrent topics or issues where disinformation and false narratives have thrived lately? Could you name a few based on your research for the briefs?

The war in Ukraine, migration, climate change, the EU, the Israel–Hamas war in Gaza and its consequences, pandemics and vaccines, and LGBTQ+ communities have all been recurring targets of disinformation narratives and campaigns in recent months (as reflected in the briefs monitoring these topics).

Moreover, national politics are frequently targeted by disinformation, although the specific dynamics naturally vary from country to country. In addition, virtually all newsworthy crises tend to become disinformation targets, at least for as long as traditional media coverage gives prominence to them (e.g. Hurricane Melissa, the theft at the Louvre, Charlie Kirk's death, the 12-day Israel–Iran war, the presidential elections in Romania, etc.).

In a recent report by the Reuters Institute for the Study of Journalism, experts forecast that verification will take centre stage in the years to come, with 'breaking verification' replacing 'breaking news'. What is your take on the evolution of fact-checking journalism and its importance in the future?

My view is that its importance will only continue to grow. We are rapidly moving toward a situation in which the main source of information for entire generations - social media - is being flooded with unreliable content, while many users are increasingly unable to distinguish what is real from what is AI-generated.

Disinformation, FIMI (foreign information and manipulation interference), scams, non-consensual AI-generated pornography (including of minors), and other illegal or harmful content and operations will thrive. These are fuelled by platforms' algorithms and business models, by unscrupulous actors exploiting the system for profit, and by adversarial/extremist forces (domestic or foreign) that benefit from polarisation and the societal crises of European states.

If democracies want to survive, they will need to address this issue decisively, and fact-checking is a fundamental tool in this effort. Even if it becomes impossible to verify all false content in the future, there should at least be a strong effort to verify what is true. Traditional sources of information could even benefit from such a shift.

Can people be taught how to detect disinformation? How can we spot a fake when we read, see or hear one? Will this even be possible given the rapid rise of AI technology, or will we again need AI to detect fakes created by AI?

A great deal can be taught. Awareness of disinformation and its main characteristics is a powerful first line of defence, and media literacy is of the utmost importance. However, education alone is not sufficient. We certainly need tools, including AI-powered ones, but currently these tools are not always reliable. Their development requires effort and investment, as bad actors are usually one step ahead. Moreover, beyond the identification of disinformation narratives, it is vital to detect and track their dissemination dynamics, including the actors, the targets and cross-platform distribution. For this type of analysis, improved AI tools can provide valuable insights for timely responses. We also need stronger regulation of the digital space and of AI. EU initiatives such as the Digital Services Act and the AI Act are a good starting point, but much more is needed, notably in terms of enforcement.

In addition, we need a strong traditional media sector capable of providing reliable, high-quality information, and more fact-checking at all levels. Above all, however, democratic governments must step up politically, boldly addressing these challenges and ensuring that the public is properly informed.

Where can people read your briefs?

You can find all of our briefs published here.

 

Tommaso Canetta is a journalist and a fact-checker, deputy director of Pagella Politica/Facta news, coordinator of the fact-checking activities of EDMO and Italian Digital Media Observatory (IDMO), and member of the Governance Body of the European Fact-Checking Standards Network (EFCSN) as well as of the Taskforce of the Code of Practice on Disinformation.

EDMO is an EU-funded network that brings together researchers, fact-checkers and media literacy experts to detect, analyse and counter disinformation across Europe. Its fact-checking network is made up of 15 hubs across the EU and the EEA.

Online disinformation is evolving fast, and Europe is facing growing threats from foreign interference and false information. Debunk.org is a disinformation analysis centre that monitors online false narratives, fact-checks claims and exposes coordinated manipulation efforts. Viktor Dauškas, head of the organisation, talked to us about their work in tracking these campaigns. He explained the techniques and patterns used in modern disinformation efforts, the challenges they face in identifying manipulation across different platforms, and the steps that institutions, civil society and individuals can take to protect democratic debate in Europe.

Based on your recent analyses (on surging pro-Kremlin narratives on X by the Telegram news channel Belarusian Silovik, deepfake scams, and a campaign to discredit Ursula von der Leyen), what are the most prominent disinformation patterns you are currently observing in Europe? How do fakes work? Which techniques are proving most effective today, and why?

The Debunk.org team observed a prominent pattern: cross-platform content laundering, which also is a new way to overcome EU sanctions on Kremlin media. Content from sanctioned websites is rewritten with AI and shared across newly created disinformation media outlets, for example Pravda network, or shared on social media channels. Narratives often originate in closed or semi-closed spaces such as Telegram, where state-linked or proxy actors operate with limited oversight, and are then repackaged for platforms like X. In our investigation into Belarusian silovik-linked content, we saw a small set of inauthentic or impersonating accounts repeatedly amplifying the same source material, using AI to localise it and tailoring different audiences to increase reach.

A second pattern is event-driven political discreditation, where formal democratic procedures are deliberately instrumentalised to launch disinformation campaigns. In the case of the no-confidence vote targeting European Commission President Ursula von der Leyen, the vote itself became a vehicle for a coordinated discreditation campaign. Although it was evident in advance that there were insufficient votes to remove the President, the process was still pursued in order to amplify pre-existing narratives of corruption and illegitimacy.

Finally, we increasingly see disinformation converging with fraud, particularly through paid advertising and deepfake-enabled scams. Here, the goal is not only political manipulation but also direct financial harm, often achieved by impersonating trusted media outlets, public figures or institutions. These campaigns benefit from platform advertising systems and are difficult for users to distinguish from legitimate content.

Can you walk us through your debunking process? What signals or evidence typically allow you to identify coordinated or manipulated content?

Our debunking process combines narrative analysis, behavioural signals and technical verification, often in collaboration with partner organisations.

We start by identifying the upstream source of a claim and mapping how it spreads across platforms. Coordination signals include unusually high posting frequency, synchronised amplification around key events, repeated linking to the same origin, and accounts that impersonate real people or organisations. At the content level, we look for recurring phrasing and templated storytelling.

We assess scale and velocity. When large volumes of content appear in short timeframes, this might suggest automation or coordinated behaviour. While tools and techniques vary by case, the key is triangulating multiple indicators of manipulation rather than relying on a single signal.

Foreign information manipulation and interference (FIMI) analysis today is rarely a solo effort. Effective responses increasingly depend on information-sharing networks like FIMI-ISAC that connect researchers, journalists, civil society and public institutions, allowing threats to be identified and contextualised quickly.

Your analysis of disinformation targeting Ursula von der Leyen highlights broader trends in political discreditation. What does this case reveal about how EU-level institutions or figures are framed online today?

This reflects a wider FIMI strategy: eroding trust and weaken EU institutional authority so that official information, media, fact-checking or policy responses are more easily dismissed. The repetition of pre-existing made up narratives or accusations across Kremlin-aligned outlets, channels and influencers attempts to create large-scale campaigns to manipulate public opinion.

Based on your research, what skills do citizens most urgently need today to protect themselves from disinformation and deepfake content? Where do you see the biggest gaps in public digital/media literacy?

The most urgent skills today relate less to fact memorisation and more to recognising manipulation techniques.

Citizens need strong source-checking habits, including lateral reading, which involves leaving a website to check what other credible sources say, and basic verification of who is behind a claim. There is also a growing need for synthetic-media awareness: understanding that a convincing video or audio can be fabricated, and that visual realism is no longer proof of authenticity.

To help address these gaps, Debunk.org developed InfoShield, a free 45-minute online course that has already been completed by more than 5 000 citizens. The course focuses on practical, everyday skills for recognising manipulation, emotional framing and deceptive content in digital environments.

At the professional level, there is also a growing need for trained specialists who can systematically identify and respond to foreign information manipulation and interference. For this purpose, we offer FIMI101, a professional e-learning course designed to certify analysts working in this field. Participation in these professional courses directly supports Debunk.org’s continued research, monitoring and public-interest activities.

Viktoras Daukšas has been at the helm of the independent technology think tank and NGO Debunk.org for eight years. Debunk.org carries out analyses of foreign information manipulation and interference (FIMI) and coordinated inauthentic behaviour analyses (CIB). Together with its partners, it researches disinformation in 22 countries by combining expert knowledge with AI-driven technologies. Debunk.org also offers educational media literacy campaigns to teach people how to spot fakes online and shield themselves from disinformation.

Europe’s hospitals faced nearly 300 cybersecurity incidents in 2024, making healthcare the most targeted essential sector. Widely attributed to Russian-linked groups, major incidents cost around EUR 300 000 each — but the damage goes far beyond financial losses. The European Commission’s 2025 Action Plan on the Cybersecurity of Hospitals and Healthcare Providers is a critical step towards protecting EU healthcare from hybrid threats. Samuel Goodger and Elizabeth Kuiper of the European Policy Centre outline the priorities for ensuring the plan’s successful implementation.

The growing number of cyberattacks against the EU’s health infrastructure form part of broader hybrid warfare intended to intimidate, destabilise and test European resolve, chiefly led by Russia. As digital health and artificial intelligence reshape healthcare provision, the cyberattack surface expands in tandem. Since 2023, pro-Russia hacker groups – such as Killnet and Anonymous Sudan – have launched coordinated attacks on hospitals and health authorities in Denmark, the Netherlands, Spain and Sweden. In 2024 alone, at least 289 cybersecurity incidents affected EU healthcare providers – more than in any other essential sector.

The cost of inaction is staggering. Major incidents cost an average of EUR 300 000 each, meaning the cumulative burden on health systems may reach billions.

Disinformation, for instance shared on social media, can also multiply attacks’ impact. When hospitals are targeted, false claims about patient data breaches can amplify public anxiety, erode trust in healthcare institutions and compound the already-concerning effects of low health literacy.

Why healthcare?

Several factors make health systems attractive targets. Personal health records enable identity theft or extortion. Fragmented IT environments – legacy systems alongside modern infrastructure – contribute to vulnerabilities. Supply-chain dependencies create additional entry points, as one system’s breach can cascade into others.

Cybersecurity preparedness in healthcare varies dramatically across the EU. While some Member States have mature ecosystems – such as the Dutch Z-CERT, which provides sector-specific threat intelligence and incident response – others lack health-specific expertise. This fragmentation creates vulnerabilities that hostile actors can exploit. Limited cross-border threat-intelligence sharing allows attackers to reuse the same vulnerabilities across countries.

Workforce shortages also exacerbate such gaps: in 2024, the EU lacked an estimated 300 000 cybersecurity professionals. The problem is particularly acute in healthcare, where roughly two-thirds of cybersecurity roles are filled by non-specialist IT professionals.

AI - A pivotal opportunity

In these circumstances, the Commission’s January 2025 Action Plan on the Cybersecurity of Hospitals and Healthcare Providers is a critical step forward. Building on substantial existing legislation – such as NIS2, GDPR and the European Health Data Space Regulation – the Plan charts a path to protect EU health systems through four pillars: Prevent, Detect, Respond and Recover, and Deter.

Today’s AI-based tools offer significant defensive potential: continuous surveillance, subtle compromise detection, alert prioritisation and automated early threat containment. However, adversaries also benefit from such evolutions – for instance by manipulating AI models with adversarial inputs or data poisoning. Ensuring system integrity therefore requires continuous monitoring and secure development pipelines. Validation by human analysts remains crucial for accountability.

AI also significantly strengthens disinformation actors. By analysing stolen data, attackers can generate phishing emails tailored to individuals’ specific roles. During incidents, coordinated disinformation can erode public confidence precisely when trust is most fragile.

Recommendations

Prior to further action by the Commission to implement the action plan, we identify six priorities:

First, leverage AI for threat detection and response. Health systems should pilot specialised AI for automated vulnerability management and behavioural analysis. Closed but explainable AI systems are preferable, to reduce data leakage risks.

Second, enhance cross-border threat intelligence. The Commission must establish vulnerability watch systems contextualised in clinical workflows. International cooperation should be strengthened through the International Counter Ransomware Initiative and G7.

Third, strengthen joint procurement for supply-chain security. Establishing common procurement mechanisms at EU level would aggregate demand and facilitate oversight of secure-by-design requirements.

Fourth, address the workforce capacity crisis. Healthcare workers themselves are both the first line of defence and a key vulnerability. Cyber hygiene training must include counter-disinformation skills and recognition of AI-enhanced social engineering attacks.

Fifth, target disinformation risks. The Commission should develop healthcare-specific AI literacy initiatives explaining decision-making processes and privacy implications. Citizens must be empowered to distinguish genuine communications from manipulated content.

Sixth, ensure adequate funding is available. In addition to redirecting existing resources, public investments should qualify under the Stability and Growth Pact’s escape clause. The EU should explore creating a dedicated EUR 10 billion Resilience Fund for sectors most exposed to cyber threats.

Ensuring the cyber resilience of EU health systems requires a shift towards a collaborative, proactive approach. This means moving beyond fragmentation and favouring integrated, innovative collective action. By capitalising on AI, deepening cross-border cooperation, investing in workforce development and empowering patients, the EU can transform the healthcare sector from a vulnerable target to resilient infrastructure.

Samuel Goodger is Policy Analyst, and Elizabeth Kuiper is Associate Director at the European Policy Centre. This article draws on their November 2025 Policy Brief 'From ransomware to statecraft: Protecting EU healthcare in the new threat landscape'.

The European Policy Centre (EPC) is an independent, not-for-profit think tank dedicated to fostering European integration through analysis and debate, supporting and challenging decision makers at all levels to make informed decisions based on evidence and analysis, and providing a platform to engage partners, stakeholders and individuals in EU policy making and in the debate about the future of Europe.

Designers and developers of digital interfaces should be more aware of people’s varying information needs. Bart Simons of the European Blind Union (EBU) urges them to step into the shoes of people with disabilities ─ even if only for a moment ─ to make the digital revolution benefit us all.

Access to information is essential. Blind and partially sighted people have long been pioneers in developing technical tools to consult printed information, as this is crucial for an independent life. We were among the first to use scanners to read text from paper and even ten years ago we already had AI tools on our smart phones to describe our surroundings.

We are generally grateful that information and many processes are becoming available digitally. However, digitalisation must be implemented in a smart and inclusive way. There is great potential for designing and developing websites, apps, banking services, books and shopping platforms that are accessible to users with widely diverse needs. Legislation such as the European Accessibility Act is in place, standards have been developed and smartphones and computers can be personalised and equipped with assistive technologies so that everyone can use them.

However, we need more awareness among designers and developers of digital interfaces. They need to be trained so that the potential of digitalisation to meet our information-access needs becomes a reality.

We also want to check the amount we are paying before pressing the OK button on the payment terminal. We cannot drive cars yet so we rely on home delivery of groceries, but that only works if the shop's website can be used without a mouse. We want to read books released yesterday and find accurate information on the internet, but that requires sufficient colour contrast. In addition, information needs to be provided in text rather than just images.

Everyone involved in creating products and services with a digital interface can help unlock this potential by looking at things from different perspectives: how do I find the double espresso button on a touch screen when I close my eyes or forget my glasses? Can users order food from this kiosk when they are short, tall or seated? On an e-learning platform, can users answer questions without using a mouse? Is there an alternative to drag and drop? Are exercises designed not to rely solely on colour codes, image recognition or other sensory characteristics?

When products and services are designed and developed inclusively more customers are reached, and those customers will feel more independent and recommend them to others. Let us make the digital revolution a reality for all.

Bart Simons is accessibility expert and representative of the European Blind Union (EBU) at the European Consumer Voice in Standardisation (ANEC).

Although AI could make European agriculture more competitive and sustainable, most farmers are being left out: under the current CAP, only around 3-4% of EU farms are expected to receive funding for digital tools, writes Stoyan Tchoukanov, president of the EESC's section for Agriculture, Rural Development and the Environment.

At the EU AgriFoodDays in December 2025, an important reality became clear. Out of more than 400 participants discussing the future of digital farming, only six were farmers and only two were digitally connected to their farms. This underlines a critical gap: Europe cannot design a digital future for agriculture without farmers being actively involved.

Although digitalisation has been discussed for decades, real support on the ground remains limited. Under the current EU Common Agricultural Policy (CAP) for 2023-2027, only about 3-4% of EU farms are expected to receive funding specifically for digital farming technologies. Even when broader farm modernisation measures are included, the figure reaches only around 7.6%.

There is also a clear divide between basic digital use and real digital transformation. While more than 90% of farmers use at least one IT or software tool, far fewer can afford or access advanced technologies such as precision farming, sensors, or AI-based systems. Dedicated public support for these investments remains low.

This is reflected in the data. According to the EU Joint Research Centre, only about one in five farmers uses farm-management software or satellite-based tools, and drone use remains minimal across most Member States.

Digital tools and AI can make European agriculture more competitive, sustainable, and resilient, but only if farmers are connected, supported, and involved in shaping these solutions. Without them, digital innovation risks remaining a policy ambition rather than a reality in the field.

Stoyan Tchoukanov has been president of the EESC's NAT section since October 2025 and an EESC member since 2020. Within the EESC, he represents the Beef Breeders Association of Bulgaria, where he also runs his own cattle farm.

Anastasia Karagianni from VUB (Vrije Universiteit Brussel) explores how digital technologies increasingly influence how people are judged and treated, from online images to access to jobs and public services. Although these systems are often presented as neutral, they can reinforce existing inequalities and cause real harm to marginalised communities, showing why EU digital regulation must move beyond technical compliance and take people’s lived experiences seriously when addressing algorithmic discrimination.

Algorithmic discrimination refers to automated systems producing outcomes that systematically disadvantage particular groups, not due to technical 'errors' alone but because of how data, design choices, and historical patterns of inequality shape machine decision‑making. These effects are especially pressing where gender, race, class, disability, or other identity axes intersect, undermining equality, privacy, and non‑discrimination.

For example, beauty filters encode normative, often Eurocentric and gendered ideals of attractiveness by algorithmically 'correcting' faces toward lighter skin tones or feminised features, disproportionately affecting women and people of colour and reinforcing existing hierarchies of social value. Similarly, smart wearable technologies, such as Ray-Ban Meta AI glasses, raise concerns about surveillance, privacy, and image-based sexual abuse, as biased vision and speech systems can misidentify marginalised groups and expose bystanders to recording without their consent, reinforcing existing power imbalances in public spaces.

In the EU, where digital systems increasingly determine access to public services, employment opportunities, and social support, addressing these harms is central to protecting fundamental rights and democratic accountability.

EU frameworks, such as the General Data Protection Regulation (GDPR) and the EU Artificial Intelligence Act (AI Act) represent important steps towards a rights‑based approach to data and automated systems. The GDPR’s emphasis on transparency, human oversight, and mechanisms for individuals to contest automated decisions gives civil society tools to challenge discriminatory practices and to demand accountability from both private and public actors. The AI Act adopts a risk-based approach to regulating AI, with explicit obligations for high-risk systems ─ AI applications considered likely to significantly affect people’s rights, safety, or access to essential services, such as healthcare or employment. This creates avenues for oversight and structured scrutiny of technologies that could produce harmful outcomes.

Civil society organisations have played a key role in bringing these frameworks to life. Forums such as the CPDP (Computers, Privacy and Data Protection Conference), Privacy Camp, and FARI engage activists, researchers, and policy-makers in evaluating algorithmic systems and shaping best practices. Successes achieved by European Digital Rights (EDRi) and the Digital Freedom Fund (DFF) demonstrate how sustained civil engagement can improve transparency obligations, strengthen enforcement, and widen public understanding of digital harms. These initiatives show that EU regulation can empower civil society, fostering participatory approaches to regulation rather than leaving oversight solely to state or corporate actors.

Despite these positive developments, significant gaps remain that limit the capacity of EU regulation to address structural discrimination and algorithmic harm in a comprehensive way. At the heart of this critique is the nature of the AI Act’s risk classification system. The Act’s reliance on a top‑down model, where regulators pre‑define categories of high‑risk systems, leaves little space for bottom‑up identification of emerging harms discovered through lived experience or civil society monitoring. Once systems are deployed, there are limited mechanisms for affected communities to trigger risk reassessments or demand remediation outside predefined categories.

The Digital Omnibus Proposal illustrates another worrying trend. By allowing providers of AI systems to self‑register and determine whether their technology qualifies as high‑risk, the proposal risks delegating critical regulatory judgments to the very actors whose commercial interests may conflict with public safety and rights protection.

Even where bias-mitigation obligations (efforts designed to reduce discrimination in AI systems) exist, they often require the processing of sensitive data. Yet gender and LGBTQIA+ characteristics, such as non-binary, transgender, or intersex identities, are frequently not recognised as protected categories and therefore remain insufficiently safeguarded. This creates blind spots in understanding how AI systems can reinforce overlapping forms of discrimination.

These gaps become most apparent  with emerging harms, such as sexualised deepfakes. While it is likely that such technologies could fall under Article 5’s prohibited practices, the regulatory text leaves ambiguity around classification and enforcement. In the absence of clear obligations on platforms to prevent or remediate image‑based abuse and deepfake dissemination, victims may find limited legal recourse, despite substantive harms to privacy, dignity, and safety.

Another limitation lies in standardisation obligations, which apply only to high‑risk AI systems. This leaves vast swathes of widely deployed technologies, including generative AI and content moderation applications, without systematic safety, fairness, and discrimination safeguards. For civil society, this means that many discriminatory or harmful systems may never be subject to robust conformity assessments or accountability pathways.

Finally, the way EU law handles intersectionality ─ the idea that people can face overlapping forms of discrimination ─ shows that current regulations don’t always reflect people's lived realities. While the Directive on Combating Violence Against Women and Domestic Violence (GBV Directive) introduces the concept of 'intersectional discrimination', its practical scope remains limited in the text of the (GBV) Directive. It also does not fully account for the concerns of LGBTQIA+ communities across EU equality policy. Academic analysis of the AI Act shows that references to 'gender equality' are sparse, and inclusive terminology for diverse gender identities is largely missing. As a result, the regulatory framework still remains rooted in binary understandings of gender.

These critiques point to a broader issue: simply following procedural safeguards is not enough to tackle algorithmic discrimination in society. What is needed are approaches that start from people's experiences and identify harms early, assessments that consider how different forms of discrimination overlap, and participatory oversight that meaningfully includes civil society in decision-making. Tools such as gender‑responsive impact assessments and community‑driven evaluation frameworks ─ which involve testing systems for bias and listening to affected users ─ can help make sure that regulation actually protects those most vulnerable to algorithmic harms. Without such mechanisms, EU digital regulation risks enshrining a 'neutral' approach that obscures the inequalities people face in everyday life, instead of confronting them.

Anastasia Karagianni is a doctoral researcher at the Law, Science, Technology and Society (LSTS) research group of the Law and Criminology Faculty at Vrije Universiteit Brussel (VUB) and former FARI scholar. Her thesis focuses on the 'Divergencies of Gender Discrimination in the EU AI Act Through Feminist Epistemologies and Epistemic Controversies'. She has been a visiting researcher at the iCourts research team of the University of Copenhagen and the Joint Research Centre of the European Commission in Seville as well as a visiting lecturer at the ITACA Institute of the UPV Universitat Politèchnica de València.