Facts

• A European Parliament (EP) committee is set to finalize its landmark Artificial Intelligence (AI) Act on Thursday before sending it for discussion between member states, Parliament, and the 27-member bloc's executive Commission.
• Yet to be finalized is a proposal to ban the use of AI facial recognition technologies in public places, including matching faces with images from the internet. It would also ban using AI to identify emotional states based on facial expressions and predictive policing.
• The center-right European People's Party is set to oppose the facial recognition ban, citing potential uses for fighting crime and terrorism, with members voting on that proposal separately from the others in the act. Exceptions will be made for medical and therapeutic use of the technology.
• Rights groups have criticized the wording of the measure for potentially allowing AI to be used against migrants through location-based predictive policing, with Amnesty International arguing they would not be protected from "AI-induced harm" to the extent EU citizens will be.
• The AI Act has been two years in the making, with lawmakers working to expand its scope after the release of ChatGPT. The act, one of the first pieces of AI regulation in the world, would also require companies to disclose copyrighted materials in their large language models.
• The act is expected to be passed by 2024 at the latest before a grace period to allow companies to adapt to the new regulations. Entities could be fined up to 6% of their global revenue if found in violation of the act.