The agreement mandates models like ChatGPT and general AI to follow transparency rules before market entry.
3164 Total views
17 Total shares
The European Parliament and Council negotiators reached a provisional agreement on the rules governing the use of artificial intelligence (AI) on Friday, Dec 8.
The agreement covers the governmental use of AI in biometric surveillance, how to regulate AI systems such as ChatGPT, and the transparency rules to follow before market entry. This covers technical documents, adherence to European Union copyright and sharing training content summaries.
The EU wants to be the first supranational authority with laws on AI, specifying how it is to be used beneficially while protecting against risks. The deal was struck following a nearly 24-hour debate on Dec. 8 and 15 hours of negotiations after that.
The agreement stipulates that AI models with significant impact and systemic risks must evaluate and address those risks, perform adversarial testing for system resilience, report incidents to the European Commission, ensure cybersecurity and disclose energy efficiency.
“Correct implementation will be essential – the Parliament will keep a close eye on supporting new business ideas with sandboxes and effective rules for the most powerful models.”
After the deal was made, European Commissioner for Internal Market Thierry Breton posted on X (formerly Twitter), “Historic! The #AIAct is much more than a rulebook — it’s a launchpad for EU startups and researchers to lead the global AI race. The best is yet to come!”
The EU becomes the very first continent to set clear rules for the use of AI
The #AIAct is much more than a rulebook — it’s a launchpad for EU startups and researchers to lead the global AI race.
The best is yet to come! pic.twitter.com/W9rths31MU
— Thierry Breton (@ThierryBreton) December 8, 2023
According to the agreement, general-purpose artificial intelligence with risks must follow codes. Governments can only use real-time biometric surveillance in specific cases like certain crimes or severe threats in public spaces.
The deal forbids cognitive behavioral manipulation, scraping facial images from the internet or CCTV footage, social scoring and biometric systems inferring personal details like beliefs and orientation. Consumers would have the right to file complaints and get explanations.
Fines for violations would range from 7.5 million euros ($8.1 million) or 1.5% of turnover up to 35 million euros ($37.7 million) or 7% of global turnover, depending on the infringement and size of the company.
According to the statement by the European Parliament, the agreed text will now have to be formally adopted by the parliament and council before becoming EU law. The parliament’s internal market and civil liberties committees will vote on the agreement at an upcoming meeting.