Accountable Tech Statement on Advancement of California’s Landmark AI Safety Bill

Today, the California General Assembly passed the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047) — a landmark bill that would require AI developers to adhere to common sense safety guidelines for new AI models. The legislation, which is sponsored by California State Senator Scott Wiener, would introduce crucial guardrails around AI development, including requiring pre-deployment safety testing and third-party auditing. The bill will head back to the State Senate for a confirmation vote and then to Governor Gavin Newsom’s desk.

Accountable Tech Co-Founder and Executive Director Nicole Gill released the following statement:

“Today, lawmakers in California are once again leading the nation on tech safety with the passage of SB 1047. This bill marks a significant move toward safer AI development and is particularly urgent absent federal regulation.

“In the last year we’ve seen an explosion of rapidly deployed AI models that supercharge the spread of disinformation and harmful content. While Big Tech companies have made empty pledges to protect our elections, their AI products are continuously used to develop deceptive deepfakes, impersonate political candidates, and create disinformation that undermines faith in our democratic institutions. We can’t afford to keep betting our safety – and democracy – on voluntary commitments.

“SB 1047 will finally force Big Tech developers to prioritize our safety over their race for profits. We urge Governor Newsom to sign the bill immediately and urge Congress to follow California’s lead with bold, decisive action to ensure responsible AI development.”

Background on Accountable Tech’s work to rein in AI:

Ahead of the Assembly’s final vote, Accountable Tech launched a campaign encouraging California residents to call their Assemblymember and ask them to support SB 1047 — an effort that led to dozens of constituent calls.

Earlier this month, Accountable Tech joined 40 civil society organizations to write a joint letter demanding Big Tech implement safeguards against the rampant spread of political deepfakes.

In 2023, Accountable Tech launched digital ads and a report detailing the broken promises made in Big Tech’s “Accord” to address AI election content signed by companies including Microsoft, Google, and OpenAI.

Learn more about Accountable Tech’s previous work on AI HERE.

Statements

Aug 16, 2024

Accountable Tech on Appeals Court Ruling on the California Age-Appropriate Design Code Act

Accountable Tech Executive Director and Co-Founder Nicole Gill issued the following statement on the Ninth Circuit Court’s narrowing of the district court’s injunction blocking California’s Age-Appropriate Design Code: The Ninth Circuit’s ruling makes clear that the lower court’s injunction was far too broad. The district court must now reconsider the safety-by-design and privacy-by-default protections central to the Age-Appropriate Design Code.

Statements

Aug 12, 2024

New: Meta’s Political Content Limit Causes 65% Drop in Reach for Prominent Accounts

Accountable Tech released a new report today that assessed the impact of Meta’s political content limit setting that the company announced in February. Researchers found that the implementation of this policy caused a 65% decrease in reach for the average piece of content posted from the prominent Instagram accounts studied.