Press release from swissAI dated January 20, 2026

With the help of AI, it is now possible to create or alter images of real people that look deceptively real. This technology is increasingly being misused to create sexualized deepfakes. In these, the faces of real people are inserted into pornographic or compromising depictions without their knowledge. The images appear authentic, spread rapidly, and are nearly impossible to remove completely from the internet.
For those affected, this is not a minor technical issue, but digital violence—with massive psychological, social, and professional consequences. Women and minors are particularly affected.

Why This Problem Exists
This phenomenon arises from a problematic combination:

1. Extremely powerful and easily accessible AI tools
What used to require specialized knowledge is now available to every user.

2. Safety measures that aren't keeping up
Many platforms release new AI features before it is clear how to effectively prevent misuse.

3. Responses Only After Harm Has Occurred
Abuse is often taken seriously only after cases become public—not as a preventive measure.

In short: Technology is advancing faster than our ability to take responsibility for it.

Current Debate Over Grok/X in the United Kingdom
In the United Kingdom, this issue is currently at the center of a political controversy. The regulatory authority Ofcom is investigating the X platform in connection with the AI chatbot Grok. The allegation: This AI functionality was used to create and distribute a large number of sexualized, AI-generated images. The British government has publicly stated that X may not be adequately protecting its users. Under the UK Online Safety Act, severe sanctions—including the temporary suspension of the platform—are on the table. X responded by restricting certain image features to paying Grok users only.
This measure falls short: it does not prevent abuse, but merely shifts it behind a paywall. Meanwhile, after weeks of pressure from regulatory authorities, governments, and victims, Elon Musk himself announced that Grok needs a “code of conduct.”

Not an isolated incident, but a structural pattern
Grok/X exemplifies a recurring pattern: New AI features are introduced, abuse occurs, and attempts to correct the situation follow in response to public pressure. For those affected, the damage remains.

swissAI's Position
The core problem lies not with AI itself, but with the release of features that are particularly prone to abuse without effective oversight.

Three points are key:

1. Not every AI function is justifiable
If a function systematically results in harming or degrading people, its very existence must be called into question.

2. Retrospective measures do not provide protection
Those who wait until abuse becomes public to take action are reacting too late.

3. Paywalls Are Not a Solution
Harmful or illegal content does not disappear simply because it becomes paywalled

swissAI calls for neither blanket bans on AI nor detailed regulatory blueprints. However, we believe it is necessary to have an open discussion about whether particularly dangerous AI functionalities should be temporarily disabled if platforms are unable to control them effectively

Why This Debate Must Take Place Now
Most people don't even realize yet what is technically possible. That is precisely why we need an objective, understandable, and forward-looking assessment now—before the damage spreads further.

Innovation is important. However, protecting people is not an optional side issue, but a fundamental requirement.