Euractiv has reported that members of the European Parliament (MEPs) have finally bridged their differences and reached a provisional political deal on the AI Act.

The text might still be subject to minor adjustments at the technical level ahead of a key committee vote scheduled on 11 May, but it is expected to go to a plenary vote in mid-June.

The last update from Parliament related to an 11th hour cancellation of the vote that is now due to take place in a few weeks. It seems that MEPs only needed a few more days to resolve their differences which appeared to be substantial at the time.

The pace of progression of this legislation and the collective energy to coalesce around some very challenging issues demonstrates the significant appetite at EU level to get this legislation over the line as soon as possible.

The final changes include:

  • The threshold for classification as an Annex III high-risk AI systems will be raised
  • Recommender systems of very large online platforms, as defined under the Digital Services Act (DSA), will be deemed high-risk
  • Strict controls on the foundation models that sit behind General Purpose AI, like ChatGPT
  • A carve out from prohibited AI for general monitoring of interpersonal communications
  • A ban on the use of emotion recognition AI-powered software in the areas of law enforcement, border management, workplace, and education
  • An extension of the ban on real-time biometric ID software to retrospective use also - although this could be challenged again

Following months of intense negotiations, members of the European Parliament (MEPs) have bridged their difference and reached a provisional political deal on the world’s first Artificial Intelligence rulebook.