AM-44-1 to R4: European Digital Democracy: Sovereignty, Rights and Online Safety

1. Digital sovereignty and transparency.

Europe’s digital infrastructures and social platforms are dominated by Google,
Apple, Meta, Amazon and Microsoft. TikTok, operated from China, handles the data
of over 125 million EU citizens, while Chinese AI systems export models based on
surveillance and censorship. With 90% of European data processed abroad, this
dependence undermines our sovereignty, security and democracy.

To regain control, the EU must act not only through regulation but also through
its market power. Public procurement should be used as a lever to support the
adoption and scaling of open and interoperable European alternatives to Big
Tech. At the same time, social media infrastructures must be opened up to third-
party users, preventing monopolies and limiting the addictive potential of
algorithms that weaken pluralism. Finally, Big Tech lobbying must be strongly
supervised and radically transparent, with an obligation to disclose which
political parties are targeting citizens with paid content on their platforms.

2. Human impact.

AI-driven manipulation has direct consequences on people’s lives.

Studies show that 96% of deepfakes are pornographic, with over 90% targeting
women. Such content, which circulates widely on platforms such as Pornhub, X, or
Telegram, fuels harassment, blackmail, and trauma.

Children and young people are particularly exposed: European teenagers spend an
average of 3 hours per day on social media (EU Kids Online 2022), where
addictive feeds increase risks of scams, explicit and violent content and
disinformation. Current recommender systems based on behavioural analysis make
the problem even worse. According to a 2023 Amnesty International report, after
only a few hours on TikTok, test accounts created as teenagers were quickly
shown harmful content — within 3 to 20 minutes, more than half of the
recommended videos were about self-harm, depression, or suicide.

Beyond children, vulnerable groups are being deliberately targeted. Meta was
fined €390 million in 2023 by the Irish Data Protection Commission for forcing
users into behavioural advertising without consent, a clear violation of the
GDPR. Later that year, the European Data Protection Board imposed a record €1.2
billion fine for illegal data transfers to the United States.

Investigations have also revealed how Meta and Google secretly partnered to
target teenagers (13–17 years old) with ads promoting Instagram on YouTube,
bypassing EU restrictions on advertising to minors. Meanwhile, studies show that
73% of European Facebook users are classified under sensitive categories such as
health, religion, or sexual orientation for ad targeting.

Besides being shown harmful content on social media platforms, users can also be shown this kind of content through chatbots and other generative AI tools. Conversations with chatbots are even less visible to parents and peers than content shown on social media, and users are at risk of emotional dependence on the tool. In April 2025, there was the case of Adam in California, who committed suicide after several conversations with the AI-tool ChatGPT. The tool did not steer the teenager away from the topic of suicide, nor did it refer to crisis services. OpenAI has since tightened ChatGPT's guardrails, only to loosen them again a few months later.

These practices show that the threat is not just about who owns the platforms,
but about how the system itself operates, exploiting emotions, vulnerabilities,
and even mental health for profit.

3. Environmental and global impact.

Digital technologies also have a heavy ecological footprint. Data centres
already consume around 3% of EU electricity and could reach 10% by 2030 (EEA).
Training large AI models and expanding cloud infrastructures will further
increase emissions unless strict standards are enforced. A sovereign digital
transition must also be a sustainable one, aligned with the Green Deal and the
1.5°C pathway.

At the same time, the EU must address the global dimension of behavioural
analysis technologies. From mass surveillance in authoritarian regimes to
political manipulation from social media platforms and exploitative mining of
rare materials, Europe’s choices have global consequences. Sovereignty cannot
mean isolation: it must mean responsibility. A Green Europe must set standards
that protect democracy, human rights and the environment worldwide.