7 of the top companies in the artificial intelligence space will meet with the Biden administration on Friday.
Leaders from Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI will meet with President Biden to agree to certain commitments that will manage possible risks from the rapidly growing technology, The Verge reports.
There are two topics that will be at the forefront of this meeting: third-party testing and watermarking. Companies are expected to commit to conducting third-party testing prior to releasing an AI product as well as developing and implementing a watermark system so that consumers know when a piece of audio or video has been developed by the software. A White House official told Gizmodo that a major reason for the watermarking system is to fight the rise of AI deep fakes — relevant given that Ron DeSantis just used AI to create a Donald Trump voice for a political ad.
The companies will also agree to share their findings on safety and attempts to bypass these safeguards. Speaking of safety, the seven companies will invest in more cybersecurity and measures to protect their data against insider threats. The AI heads have agreed to prioritize research “showing the risks of bias, privacy, and discrimination” in their AI products.
All of these commitments will be voluntary, meaning that these companies will be policing themselves.
“We’re proud to join with other leading AI companies to jointly commit to advancing responsible practices in the development of artificial intelligence,” Kent Walker, president of global affairs for Google and Alphabet, wrote in the company’s blog. “Today is a milestone in bringing the industry together to ensure that AI helps everyone. These commitments will support efforts by the G7, the OECD, and national governments to maximize AI’s benefits and minimize its risks.
These steps aren’t new for at least a couple of the firms that will be meeting with Biden. For example, Microsoft has already pledged to watermark its AI images and videos, and OpenAI already “red teamed” GPT-4 early this year, a practice by which researchers prompt an AI software to acts unintended ways so that developers can fix these bugs prior to a public launch.
The Biden administration’s meeting with these companies further underlines the rising threat of AI. As it pertains to the entertainment industry, the fear of being replaced by artificial intelligence has been a massive concern in both the WGA strike and the SAG-AFTRA strike. Whereas the WGA is fearful that studios and streamers will use the software to write scripts, SAG-AFTRA’s concerns are slightly more abstract as the negotiations revealed the studios desire to use background actors’ likeness in perpetuity while only paying them a day rate.
