Listen to the Podcast:
The EU has made significant progress toward establishing one of the first artificial intelligence laws in the world after its main legislative body adopted the text of a draft rule that forbids police from using live face recognition technology in public areas.
The European Parliament passed regulations to create a global standard for the technology, which includes anything from deepfakes—AI-generated videos—and ChatGPT-like bots to some types of drones and automated medical diagnostics.
Before the proposed rules, called the AI act, become law, MEPs will now work out the details with EU countries.
“AI brings up a lot of questions about society, ethics, and the economy. But now is not the time to put anything on hold. “On the contrary, we need to move quickly and take responsibility,” said Thierry Breton, the European commissioner for the internal market.
A revolt by center-right MEPs in the EPP political grouping over a ban on real-time facial recognition on the streets of Europe didn’t happen because many lawmakers were in Italy for Silvio Berlusconi’s funeral.
In the end, 499 people voted yes, 28 people voted no, and 93 people didn’t vote at all.
European leaders aren’t likely to agree to a total ban on biometrics, since police all over the continent want to use the technology to spot crooks as they walk down the street or through public places.
The EPP had said that the technology could be very important for fighting crime, getting information about terrorists, and looking for lost children.
Emotional detection, which is used in some parts of China to tell if a truck driver is tired, for example, will also be against the law at work and in schools.
Roberta Metsola, president of the European Parliament, called it “legislation that will undoubtedly set the global standard for years to come.” She said that the EU could now set the tone for the whole world and that “a new age of scrutiny” had started.
Brando Benifei, a co-chair of the AI committee of the parliament, which moved the bill to the voting stage, said that the law would provide “a clear safeguard to avoid any risk of mass surveillance” when it came to face recognition.
Dragos Tudorache, one of his co-rappers, said that if the law had already been in place, the French government would not have been able to pass a law this year to allow live face recognition for crowd surveillance at the 2024 Olympics.
To stop the high risk of copyright violations, the law will require developers of AI chatbots to share all the works of scientists, musicians, illustrators, photographers, and journalists that were used to train them. Also, they’ll have to show that everything they did to train the machine was legal.
If they don’t, they could be forced to delete their apps right away or be fined up to 7% of their income, which for tech giants could be hundreds of millions of euros. “There are plenty of sharp teeth in there,” said Tudorache.
He said that talks with the European Council and the European Commission would start, and that he would go into them with a mandate from the parliament instead of “red lines” on the controversial issue of face recognition.
Benifei said that the EPP’s efforts to get rid of the blanket ban on mass surveillance because it would stop police from using the tool for security were just propaganda, since authorities would still be able to use biometric data and CCTV footage to find criminals, just like they do now.
People on both sides of the Atlantic are calling for AI to be regulated, and western governments are under more and more pressure to act quickly in what some people call a fight to save humanity.
Some people are excited about how artificial intelligence (AI) will change work, healthcare, and creative activities, but others are afraid that it could hurt democracy.
Even if the EU meets its ambitious goal of agreeing on the law by the end of the year, it won’t go into effect until at least 2026. This means that the EU has to try to get tech companies to sign a voluntary deal until then.
Margrethe Vestager, who is in charge of antitrust issues in the EU, told reporters that a balance could still be reached because the parliament has both people who support a ban because of privacy principles and people who take a “slightly more pragmatic or security-oriented approach.”