Poland needs updated laws to fight AI-driven identity theft and deepfakes

Poland faces a growing challenge in combating identity theft and the misuse of artificial intelligence, especially concerning deepfakes. While upcoming regulations like the AI Act will require labeling AI-generated content, cybercriminals are unlikely to comply, rendering such measures insufficient against malicious actors.

Experts argue that current Polish law, specifically Article 190a § 2 of the Penal Code concerning identity theft, does not fully address the nuances of generating and using misleading images. The law often requires a direct "impersonation" and proof of financial or personal harm, which may not apply to all deceptive uses of a person's likeness. There is a call for criminal liability for producing, distributing, and publicly presenting images that could mislead viewers about their authenticity, regardless of whether AI was used.

The proposed legal changes would broaden the scope to include audio, images, and video content that falsely represent an individual. This could involve altering backgrounds, adding people to different settings, or manipulating objects a person is depicted holding. The aim is to create a clear legal framework that allows for the swift removal of harmful advertisements and content from online platforms.

Implementing clearer criminal provisions would also empower individuals whose likeness is misused, granting them victim status in criminal proceedings. This would create a stronger incentive for platforms to quickly remove infringing content. The urgency is particularly high given the sensitive issue of generating non-consensual intimate imagery, including that of children, often by peers. Full protection against deepfakes is deemed essential, and mandatory labeling of AI content is seen as insufficient on its own.