Source: AFP
EU member states and lawmakers reached an agreement on Friday on how to draft “historic” rules governing artificial intelligence models like ChatGPT — after 36 hours of negotiations.
Meeting in Brussels, negotiators hammered out restrictions on how artificial intelligence can be used in Europe, which they said would not harm innovation in the field or the prospects for future European AI champions.
“Historic! With the political agreement on the AI law sealed today, the EU becomes the first continent to set clear rules for the use of artificial intelligence,” said EU Internal Market Commissioner Thierry Breton.
“The AI Act is much more than a rulebook — it’s a starting point for EU start-ups and researchers to lead the global fight for trustworthy AI,” he added.
The “AI Law” was rushed through the European Union’s legislative process this year after chatbot ChatGPT, a mass market portal for productive artificial intelligence, exploded in late 2022.
![](https://images.yen.com.gh/images/7a90eeee7d6b6f67.jpg?impolicy=cropped-image&imwidth=256)
![](https://images.yen.com.gh/images/7a90eeee7d6b6f67.jpg?impolicy=cropped-image&imwidth=256)
Read also
Once the bane of big tech, Vestager’s star is fading
While ChatGPT’s ability to generate articles, essays and poems was a dizzying illustration of the rapid advances in artificial intelligence, critics worry about how the technology can be misused.
The AI creation software, which also includes Google’s Bard chatbot, can quickly generate text, images and audio from simple commands in everyday language.
Other examples of genetic AI include Dall-E, Midjourney, and Stable Diffusion, which can create images in almost any style on demand.
Negotiators initially failed to agree after marathon talks that began on Wednesday lasted 22 hours and ended only with an agreement to resume talks the next day.
Exhausted negotiators resumed talks at 08:00 GMT on Friday.
There was no real deadline, but senior EU officials were desperate to secure a deal before the end of the year.
The European Commission, the EU’s executive arm, first proposed the law in 2021 to regulate artificial intelligence systems based on risk assessments of software models.
![](https://images.yen.com.gh/images/fc40037e3875ff06.jpg?impolicy=cropped-image&imwidth=256)
![](https://images.yen.com.gh/images/fc40037e3875ff06.jpg?impolicy=cropped-image&imwidth=256)
Read also
Learn to forget? How to rein in a rogue chatbot
The higher the risk to the rights or health of individuals, for example, the greater the obligations of the systems.
The law still needs to be formally approved by member states and parliament, but Friday’s political agreement was seen as the last major hurdle.
“The AI Act is a world first. A unique legal framework for the development of artificial intelligence that you can trust,” EU chief Ursula von der Leyen said in a social media post welcoming the agreement.
“And for the security and fundamental rights of people and businesses. A commitment we made in our political guidelines – and delivered. I welcome today’s political agreement.”
The EU is not alone in its concerns about artificial intelligence.
US President Joe Biden issued an executive order on AI safety standards in October, and while Europe is well on its way to the first broad law covering the field, Chinese legislation specifically regulating genetic artificial intelligence took effect in August this year.
![](https://images.yen.com.gh/images/d0ec6839f011befe.jpg?impolicy=cropped-image&imwidth=256)
![](https://images.yen.com.gh/images/d0ec6839f011befe.jpg?impolicy=cropped-image&imwidth=256)
Read also
The EU will resume negotiations on the world’s first artificial intelligence law on Friday
Penalties for violations
One of the main stumbling blocks during the negotiations was how to regulate general purpose AI systems such as ChatGPT.
Some member states feared that excessive regulation would harm the development of European champions such as Germany’s Aleph Alpha or France’s Mistral AI.
French digital technology minister Jean-Noel Barot said France would “carefully analyze the compromise” agreed and ensure it “maintains Europe’s ability to develop its own artificial intelligence technologies”.
The agreement includes a two-tiered approach, with transparency requirements for all general-purpose AI models and stricter requirements for the most powerful models.
Another sticking point was remote biometric surveillance — basically, facial recognition through camera data in public places. Governments wanted exemptions for law enforcement and national security purposes.
While the agreement prohibits real-time facial recognition, there will be a limited number of exceptions.
But not everyone was happy with the deal.
“Unfortunately, speed seems to have trumped quality, with potentially disastrous consequences for the European economy,” said Daniel Friedlaender, head of Europe at CCIA, one of the main technology lobby groups.
![](https://images.yen.com.gh/images/9fa60a5c2f75fd19.jpg?impolicy=cropped-image&imwidth=256)
![](https://images.yen.com.gh/images/9fa60a5c2f75fd19.jpg?impolicy=cropped-image&imwidth=256)
Read also
Google looks to lead the way in artificial intelligence with Gemini
“It may even end up driving out the European champions that the EU is so desperate to empower,” said CCIA Europe policy director Boniface de Champris.
The EU will be able to monitor and sanction those who break the law through a new body called the EU AI office which will be attached to the commission.
The office will have the power to impose a fine of seven percent of a company’s turnover or €35 million, whichever is greater.
Source: AFP