AI deepfakes spark democracy debate: Can tech save free speech?

Berlin clash exposes deep divides over AI threats to democracy

A heated debate in Berlin saw Federal Justice Minister Stefanie Hubig and attorney Christian Schertz clash over how to shield democracy from AI-driven threats. The discussion centred on deepfakes, disinformation, and the role of tech platforms in policing illegal content. Both agreed on the urgency of action but disagreed sharply on the methods needed to tackle the problem.

Meanwhile, the European Commission has ordered X (xAI) to retain internal documents related to its AI chatbot Grok until at least 2026, as regulators scrutinise its handling of controversial content. The UK’s Ofcom is also preparing potential sanctions against the platform under its Online Safety Act, including a possible ban.

At the Berlin event, Schertz painted a grim picture of democracy under siege. He warned that a dangerous mix of public outrage, AI-generated deepfakes, and disinformation could destabilise society. Urging decisive measures, he called the current moment 'democracy’s last bullet' and stressed that surrendering to extremism would hand power to democracy’s enemies.

To combat the threat, Schertz proposed two key steps: mandatory filters for platform operators to block illegal material and a real-name policy for online users. He argued that anonymity fuels abuse and that tech companies must take greater responsibility for content on their sites. Hubig, however, pushed back on several fronts. She defended online anonymity as a core principle, arguing that stripping it away could undermine free expression. Instead, she emphasised media literacy and European expertise as vital tools for protecting democratic values. On the question of banning the far-right AfD party, she adopted a cautious stance, stating that her ministry would await the Higher Administrative Court’s decision before taking action. Hubig also acknowledged the challenges of keeping pace with criminal innovation. While admitting the state cannot match the speed of digital threats, she pointed to her ministry’s work on a Digital Violence Protection Act. The proposed law aims to close enforcement gaps and strengthen responses to online harms. Beyond Germany, regulatory pressure on AI-driven platforms is mounting. The European Commission has instructed X (xAI) to preserve internal data linked to Grok until 2026, as part of a potential investigation into AI-generated deepfakes and explicit content. Key figures like EU digital spokesperson Thomas Regnier, Vice-President Henna Virkkunen, and Commission President Ursula von der Leyen are overseeing the process. In the UK, Ofcom is preparing to act against xAI and Grok under the Online Safety Act. The regulator’s possible measures include fines or even a full ban on X’s platform, reflecting broader concerns shared by French and British authorities about unchecked AI risks.

The Berlin debate highlighted deep divisions over how to balance security, free expression, and enforcement in the digital age. Germany’s proposed Digital Violence Protection Act and the EU’s scrutiny of X (xAI) signal a growing push for stricter oversight. With regulators in the UK and Europe also gearing up for action, the coming months will likely see tighter controls on AI-generated content and online platforms.