7.4 C
London
Monday, March 4, 2024

Artificial intelligence. This is how the European Union wants to regulate it

Must read

- Advertisement -


The European Union reached an agreement on Friday on rules governing the use of artificial intelligence, including governments’ use of artificial intelligence in biometric surveillance and how to regulate artificial intelligence systems such as ChatGPT.

The political agreement will make the EU the first major world power to pass laws regulating artificial intelligence. Friday’s agreement between EU countries and members of the European Parliament came after almost 15 hours of negotiations, which followed almost 24 hours of debate the previous day.

As Reuters notes, in the coming days both sides are expected to agree on details that may change the shape of the final legislation.

– Europe has taken a pioneering position, understanding the importance of its role as a global standard-setter. Yes, I think it is a historic day, European Commissioner Thierry Breton said at a press conference.

- Advertisement -

The agreement requires core models such as ChatGPT and general-purpose AI systems (GPAI) to meet transparency obligations before being released on the market. These include: preparing technical documentation, complying with EU copyright law and disseminating detailed summaries of the content used in training.

In the case of more advanced models, they will need to be regularly evaluated, assessing and mitigating systemic risks, conducting adversarial tests, reporting to the European Commission on serious incidents, ensuring cybersecurity and reporting on their energy efficiency.

Use of biometric surveillance

Governments can only use real-time biometric surveillance in public spaces on victims of certain crimes, to prevent real, present or foreseeable threats, such as terrorist attacks, and to search for people suspected of committing the most serious crimes.

The agreement prohibits cognitive-behavioral manipulation, untargeted removal of facial images from the Internet or CCTV footage, social scoring systems and biometric categorization to make inferences about political, religious, philosophical beliefs, sexual orientation and race.

Under the current state of the law, consumers would have the right to submit complaints and receive meaningful explanations, and fines for violations would range from EUR 7.5 million or 1.5 percent. turnover, up to EUR 35 million or 7%. global turnover.

A bitter compromise?

Business group DigitalEurope criticized the rules, calling them another burden on companies, alongside other recent regulations. – We reached an agreement, but at what cost? We fully supported a risk-based approach to the use of AI rather than the technology itself, but this last-minute attempt to regulate the underlying models has turned this upside down, said group CEO Cecilia Bonefeld-Dahl.

Privacy rights group European Digital Rights was equally critical.

“It’s hard to get excited about a bill that, for the first time in the EU, takes steps to legalize live public facial recognition across the EU,” senior policy adviser Ella Jakubowska said. “Although Parliament has fought hard to limit the damage, the overall package on biometric surveillance and profiling is modest at best,” she said.

The rules are expected to come into force early next year and should take effect two years later. As Reuters notes, EU law could become a model for other governments and an alternative to the lenient approach of the United States and temporary regulations China.

Main photo source: Shutterstock/Willyam Bradberry



Source link

More articles

- Advertisement -

Latest article