Tech News

It’s just become a big week for AI regulation

It’s a week for the government to back down on misuse Artificial intelligence.

Today the EU released a long-awaited set AI regulations, the initial draft of which was leaked last week. The regulations are extensive, with restrictions on mass surveillance and AI manipulation for people to manipulate.

Statement of intent by the U.S. Federal Trade Commission, a A short post on the blog by employee attorney Elisa Jillson on April 19, he may have more teeth in the immediate future. According to the publication, the FTC intends to go after companies that use and sell it biased algorithms.

Some companies will be scared right now, says Ryan Calo, a professor at the University of Washington who specializes in technology and law. “It’s not the only blog post,” he says. “This blog post is a very crude example of what a sea change will be.”

The EU is known for its tough line against Big Tech, but the FTC has taken a softer approach, at least in recent years. The agency aims to police unfair and unfair commercial practices. Its jurisdiction is narrow — it has no jurisdiction over government agencies, banks, or nonprofits. But it can come in when companies misrepresent the capabilities of a product they sell, that is, companies that claim facial recognition systems, predictive police algorithms or health tools they are not biased they can be in the line of fire. “Where they have power, they have tremendous power,” Calo says.

Performing actions

The FTC has not always been ready to seize that power. After being criticized for being too aggressive in the 1980s and 90s, he backed down and chose fewer fights, especially against tech companies. This seems to be changing.

In a blog post, the FTC warned vendors that claims about AI should be “true, not misleading, and protected by evidence”.

“For example, suppose an AI developer tells customers that their product will provide ‘100% unbiased hiring decisions,’ but the algorithm is built with data that is not racial or gender diverse. The result could be deception, discrimination, and FTC law enforcement actions.”

The FTC’s action has the support of the party in the Senate, where asked the commissioners yesterday what more they could do and what they had to do. “There’s wind behind Bethlehem,” Calo says.

Meanwhile, although they draw a clear line in the sand, EU AI regulations are only guidelines. As with the GDPR rules introduced in 2018, it will be up to each EU Member State to decide how to implement them. Some languages ​​are vague and open to interpretation. Take a “provision against subliminal techniques that go beyond a person’s consciousness to materially distort a person’s behavior” as it can cause psychological harm. Does that apply to social media news and targeted ads? “We can expect many lobbyists to try to explicitly reject advertising or referral systems,” says Michael Veale, a member of the faculty at University College London, who studies law and technology.

It will take years of legal challenges in the courts to remove details and definitions. “That will be after a very long process of investigation, complaint, fine, appeal, appeal and appeal to the European Court of Justice,” says Veal. “At that point the cycle will start again.” But the FTC, despite its close role, now has autonomy to act.

A major limitation common to the FTC and the European Commission is the inability of governments to control the use of harmful AI technology. EU regulations include carvings for the use of state surveillance, e.g. And the FTC just has permission to go after companies. Private vendors can intervene by preventing the sale of biased software to law enforcement agencies. But that will be difficult to implement because of the lack of rules about the secrecy surrounding these sales and when government agencies should declare them. acquiring technology.

Yet this week’s forecasts reflect a tremendous global shift in serious regulation of AI, a technology that has so far been developed and expanded with little oversight. Ethical guardians have been demanding restrictions on unfair and harmful AI practices for years.

The EU protects its regulations for the protection of AI freedoms. “Artificial intelligence must be at the service of the people, so artificial intelligence must always respect people’s rights.” said Ursula von der Leyen, President of the European Commission, in a speech before the release.

Regulation will help AI image problem. As Von der Leyen said, “We want to encourage citizens to have confidence in using it.”




Source link

Related Articles

Back to top button