‘Disappointed But Not Surprised’: Former Employees Speak Out on OpenAI’s Opposition to SB 1047

Two former OpenAI researchers who resigned this 12 months over security concerns say they’re disillusioned but not surprised by OpenAI’s decision to oppose California’s bill geared toward stopping AI disasters, SB 1047. Daniel Kokotajlo and William Saunders previously warned that OpenAI was in “irresponsible” race for domination.

“Sam Altman, our former boss, has repeatedly called for regulation of artificial intelligence,” they write in letter which was shared with Politico and which calls on California Gov. Gavin Newsom to sign the bill. “Now that real regulation is on the table, he opposes it.” The pair adds that “with proper regulation, we hope OpenAI can live up to its mission of safely building AGI.”

In response to the previous employees, an OpenAI spokesperson said the startup “strongly disagrees with the misinterpretation of our position on SB 1047” in a press release to TechCrunch. The spokesperson pointed to AI-related bills in Congress that OpenAI has supported, noting that “AI security regulations should be implemented at the federal level given their impact on national security and competitiveness.”

OpenAI competitor Anthropic has expressed support for the bill, while also raising specific concerns and asking for amendments. Several of those amendments have already been incorporated, and on Thursday, CEO Dario Amodei wrote to Newsom, saying the present bill “the benefits probably outweigh the costs,” without, nonetheless, fully supporting the bill.