The United States and Europe issue the strictest artificial intelligence rules to date

In recent weeks, the United States, the United Kingdom and the European Union have enacted their toughest regulations yet on the use and development of artificial intelligence, setting a precedent for other countries.

This month, the United States and the United Kingdom signed a memorandum of understanding that allows the two countries to cooperate in developing the testing of state-of-the-art artificial intelligence models to fulfill the commitments made at the Artificial Intelligence Security Summit last November.

The actions come after the European Parliament voted in March to adopt its first comprehensive set of rules on artificial intelligence. The landmark decision sets out a wide-ranging set of laws to regulate this explosive technology.

At the time, Brando Benifei, co-rapporteur for the full vote on the AI ​​​​Bill, said: “I think today is another historic day on our long road to AI regulation… The world’s first clear The prescribed regulations lead the way towards safe and human-centered AI development.”

The new rules aim to protect citizens from dangerous uses of artificial intelligence while exploring its unlimited potential.

Beth Noveck, a professor of experiential artificial intelligence at Northeastern University, expressed enthusiasm for the rules.

“It’s really exciting that the EU has actually adopted the world’s first…binding legal framework for artificial intelligence. However, this is not the end yet. This is really just the beginning.”

The new rules will be applied based on risk level: the higher the risk, the stricter the rules.

“This is not regulation of technology,” she said. “It is regulating the use of technology, seeking to prohibit, limit and control the most malicious uses – as well as transparency into other uses.

“So what China is doing in terms of social credit scores and surveillance of its citizens is unacceptable.”

Novick described what she called “high-risk uses” that would come under scrutiny. These include using tools in a way that may deprive people of their freedom or employment opportunities.

“There are also lower-risk uses, such as the use of spam filters, which involve the use of artificial intelligence or translation,” she said. “When your phone gives you weather information, it’s using artificial intelligence all the time; when you use Siri or Alexa, we’re going to see a lot less scrutiny of those common uses.”

But as AI experts point out, the new law simply creates a framework for a new governance model for the rapidly evolving technology.

Dragos Tudorache, co-rapporteur for the full vote on the AI ​​Bill, said: “Because AI will have impacts that we cannot measure through this bill alone, we must be very careful about how the technology evolves in the industry and prepare for the future. “

In late March, the Biden administration released its first government-wide policy to mitigate the risks of artificial intelligence while leveraging its benefits.

This follows an executive order issued by President Joe Biden last October calling on federal agencies to take the lead in better governing technology without stifling innovation.

“This landmark executive order demonstrates what we stand for: safety, security, trust, openness,” Biden said at the time, “proving once again that America’s strength lies not just in the strength of its example, but in the example of its power.”

Looking ahead, experts say the challenge will be updating rules and regulations as technology continues to evolve.