California State Assembly passes sweeping AI safety bill
VergeAug 25
Employees at Meta, OpenAI, Google Sign Letter in Support of California AI Bill
Above: Gov. Gavin Newsom (D-CA) reacts as he and a group of fellow democrats speak to reporters in the spin room following the CNN Presidential Debate between U.S. President Joe Biden and Republican presidential candidate, former U.S. President Donald Trump. Image copyright: Andrew Harnik/Staff/Getty Images News via Getty Images
The Facts
Approximately 120 current or former employees from artificial intelligence and tech companies such as Anthrophic, Google's DeepMind, Meta, and OpenAI have signed a letter in support of California's AI regulation bill SB-1047.
SB-1047, authored by Democratic California State Sen. Scott Wiener, requires companies to test their AI models and publicly disclose their safety protocols to stop model manipulation, with oversight from third-party auditors. It also grants the state attorney general the power to sue developers who do not comply.
The Spin
Pro-establishment narrative
California's AI bill seeks to castrate innovation within the tech industry. The text's naïve attempt to mitigate potential harm exposes a complete lack of understanding of the very sector politicians are attempting to restrict. While well-intended, SB-1047 is not the right solution to credible AI concerns.
Establishment-critical narrative
Having passed all stages of California's legislative process, all eyes are now on Gavin Newsom. While only time will tell if the governor will finally address the growing public concerns over increasingly sophisticated and dangerous AI models, recent history, unfortunately, shows the puppet politician will likely continue to bend the knee to Big Tech's demands.