President Joe Biden on Friday received pledges from American tech giants to reduce the risks of fraud and misinformation related to the rise of artificial intelligence, which was the subject of a summit with industry tenors at the White House the same day.
• Also read: Strike in Hollywood: Artificial Intelligence, a threat to actors?
• Also read: James Cameron says he warned society against weaponizing AI back in 1984
• Also read: World peace threatened by artificial intelligence, UN chief warns
Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI have agreed to “respect three principles that must be fundamental to artificial intelligence (AI) development, namely safety, security and trust,” the White House announced Friday morning.
The executives of these companies will be received there on Friday afternoon as fears surrounding this powerful technology mount, dangers arise for consumers (fraud) and citizens (misinformation) and there is a risk that many jobs will be lost.
Specifically, to make this technology more secure and transparent, these seven companies have pledged to test their computer programs internally and externally before launch.
They have also committed to investing in cybersecurity and sharing relevant information about their tools, including possible shortcomings, with authorities and researchers.
They must also “develop robust techniques to ensure users know when content has been generated by AI, such as a watermarking system,” the White House adds.
“This will allow AI-related creativity to thrive while reducing the risk of fraud and tricks,” the Biden administration says.
So far, the major companies involved have been reluctant to brand content created with their programs.
“It’s a complex issue,” a senior White House official admitted at a news conference.
The watermark “should work for both visual and audio content,” he explained. “It has to be technically robust, but also clearly recognizable for the user.”
Election 2024
Fake photos and complex montages (deepfakes) have been around for years, but generative AI, capable of producing text and images in everyday language upon simple request, is raising concerns that there could be a flood of fake content across the web.
These can be used to create highly credible scams or even manipulate public opinion. A particularly worrying prospect as the 2024 US presidential election approaches.
The support committee of Ron DeSantis, Donald Trump’s great rival in the Republican Party nomination race, has previously used a fake artificial intelligence-generated voice similar to that former president in a commercial.
“We must pull all the levers of the federal government to monitor (this sector, editor’s note)” and “work” with Congress to legislate, White House Chief of Staff Jeff Zients told information site Axios.
“We need legislation to have the experts we need in the federal government and to have the regulator to hold the private sector accountable for its actions (…)” he continued.
Current political tensions in Congress make new AI legislation unlikely anytime soon, but the government has said it is working on an executive order.
According to Paul Barrett, associate director of the Center for Business and Human Rights at New York University (NYU), these industry commitments obtained from the Biden administration are “an important first step.”
“But since they are not subject to sanctions for non-compliance, it is important that Congress legislates quickly,” he added.
The White House also pledged that it is working with foreign allies to create “a strong international framework to govern the development and use of AI” around the world.
The issue was the focus of May’s G7 summit in Japan, and the UK is expected to host an international AI summit next autumn.