Encode Justice's co-sponsorship of California SB 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (February–September 2024)

From February through September 2024, Encode Justice — the youth-led international AI-policy advocacy organisation founded by Sneha Revanur — co-sponsored California Senate Bill 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, authored by State Senator Scott Wiener. The bill, the first substantive US-state legislative attempt at frontier-AI safety regulation, was sponsored by Encode Justice alongside the Center for AI Safety Action Fund and Economic Security Action California, and was co-authored by Senator Roth (D-Riverside), Senator Susan Rubio (D-Baldwin Park), and Senator Stern (D-Los Angeles). Encode's working theory of change was that a youth-led civil-society coalition could combine three tracks — direct bill sponsorship and Sacramento lobbying, a public-mobilization track running op-eds and an entertainment-industry open letter, and validator endorsements from senior AI researchers — to push the bill through both chambers of the California legislature and onto the governor's desk. The campaign succeeded in producing the largest civil-society coalition any US-state frontier-AI safety bill had assembled to date, and lost the bill to Governor Gavin Newsom's 29 September 2024 veto.

The bill (February 2024): scope and substantive provisions

SB 1047 was introduced on 7 February 2024 with Encode Justice publicly named as a co-sponsor from the outset. As filed, the bill applied to so-called "covered models" — frontier AI systems trained using more than 10^26 floating-point operations and costing more than $100 million to train — and required their developers to conduct pre-deployment safety testing and red-teaming, implement cybersecurity safeguards and a full-shutdown capability, publish annual safety and security protocols, accept civil liability for severe harms enforceable by the California Attorney General, and provide whistleblower protections for laboratory employees. The bill also established CalCompute, a public computing cluster intended to give academic and startup researchers access to frontier-scale infrastructure outside the major labs. Encode Justice's policy framing treated SB 1047 as the first chance to make the basic safety practices that the largest AI developers had already publicly committed to a matter of enforceable state law rather than voluntary corporate policy.

Senate passage and the validator track (May 2024)

On 21 May 2024 the California Senate passed SB 1047 by a bipartisan 32–1 vote. Encode VP of Political Affairs Sunny Gandhi — who served as Encode's lead on the Sacramento track throughout the campaign — framed the Senate vote in the Wiener office's accompanying press release with the statement that "this legislation will help bring us closer to ensuring AI is developed responsibly, with the necessary guardrails to protect our generation's future." The same press release organised the campaign's validator track around the two most highly cited AI researchers on the public record — Geoffrey Hinton, the former Google AI lead, and Yoshua Bengio, the Mila and Université de Montréal computer scientist — both of whom endorsed the bill publicly. Hinton's quoted line that "SB 1047 takes a very sensible approach to balance those concerns" of AI risks and that "it's critical that we have legislation with real teeth to address the risks" became one of the campaign's most cited validator endorsements through summer 2024. Bengio published a supporting op-ed in Fortune and joined the Hinton endorsement in subsequent campaign materials.

The August amendments and Assembly passage (15–29 August 2024)

Through summer 2024 the bill drew increasingly intense opposition from major Silicon Valley developers — Google, Meta, OpenAI, Andreessen Horowitz, and Y Combinator — and from a portion of the Congressional Democratic delegation including Speaker Emerita Nancy Pelosi and House Science Committee Ranking Member Zoe Lofgren. On 15 August 2024, Wiener filed substantial amendments responding to industry feedback: the originally proposed Frontier Model Division was removed, pre-harm enforcement was narrowed, criminal perjury penalties were dropped, and a "reasonable care" standard was substituted for the original draft's "reasonable assurance." Anthropic CEO Dario Amodei subsequently wrote that the amended bill's "benefits likely outweigh its costs," a qualified endorsement that nonetheless represented the only major frontier-AI laboratory to break with the broader industry opposition.

On 28 August 2024 the California State Assembly passed the amended bill 48–16; on 29 August the Senate concurred 30–9, sending the bill to Newsom's desk. Sunny Gandhi's Assembly-passage statement — that "SB 1047 represents a critical step towards responsible development and proactive governance" and that "the bipartisan floor vote in both houses shows California lawmakers understand exactly what is at stake" — was carried in the Wiener office's press release alongside Nathan Calvin's framing that "we will be safer with SB 1047" and that "the vast majority of Californians support this bill."

The veto-eve mobilization: SAG-AFTRA, the AI-lab-employee letter, and Artists 4 Safe AI (September 2024)

The campaign's September 2024 mobilization combined three letter tracks targeted at the governor's signing decision. On 9 September 2024, SAG-AFTRA — represented by General Counsel Jeffrey Bennett, who framed SB 1047 as "the one bill that targets only the incredibly powerful expensive systems" — sent a letter to Newsom urging signature, the union's first formal engagement on broad AI safety beyond its entertainment-specific deepfake concerns. The same day, more than 113 current and former employees of OpenAI, Google DeepMind, Anthropic, Meta, and xAI signed a public letter in support of the bill.

The most prominent veto-eve artefact came on 25 September 2024 with the Artists 4 Safe AI open letter, an Encode-coordinated push that gathered signatures from more than 125 Hollywood actors, directors, and producers urging Newsom to sign SB 1047. Lead signatories included Jane Fonda, Mark Ruffalo, Shonda Rhimes, J.J. Abrams, Pedro Pascal, Judd Apatow, Rob Reiner, Mark Hamill, Ava DuVernay, Mahershala Ali, Joseph Gordon-Levitt, Don Cheadle, Jessica Chastain, Allison Janney, Fran Drescher, Sean Astin, Janelle Monáe, Mariska Hargitay, Alec Baldwin, Annaleigh Ashford, Alyssa Milano, Jessica Alba, Rosanna Arquette, John Cho, Jennifer Beals, and Tom Arnold. The letter's central framing — "this bill is not about protecting artists — it's about protecting everyone" — explicitly moved the entertainment industry's prior, sectoral AI advocacy into the broader frontier-AI safety frame Encode had been organising around. Stanford Daily reporting subsequently credited Encode Justice with coordinating the Hollywood track, noting Revanur's role in the organisational work behind the letter.

The veto (29 September 2024)

On 29 September 2024, Governor Newsom vetoed SB 1047. The veto message argued that the bill's size-only threshold "applies stringent standards to even the most basic functions" based on model size alone, without considering whether systems operated in high-risk environments — that the framework was "a solution that is not informed by an empirical trajectory analysis of AI systems and capabilities." Newsom paired the veto with an announcement that he would convene a working group on responsible AI development frameworks led by Stanford's Fei-Fei Li, Carnegie Endowment president Mariano-Florentino Cuéllar, and UC Berkeley's Jennifer Tour Chayes.

Sneha Revanur's reaction statement, released the day of the veto, called the decision "painful and disappointing" and a signal that "the tech lobby's chokehold is far tighter than the people's," citing polling showing 77% of Californians supported the bill. Revanur framed the veto, on record, as "an opening salvo. Next time, our movement will win. In some ways, it already has — we've catapulted the conversation about AI policy into the spotlight."

Tactics and posture

Across eight months, the campaign settled into a recognisable repertoire: formal bill co-sponsorship and direct Sacramento legislative work by Encode's policy staff (Sunny Gandhi leading on political affairs, with Nathan Calvin on state legal strategy and Seve Christian on California operations); validator endorsements from the two most-cited AI researchers on the public record (Hinton and Bengio) anchoring the technical-credibility track; a coalition-letter cadence escalating from May Senate-floor statements through the August Anthropic qualified endorsement, the September 9 SAG-AFTRA and AI-lab-employee letters, and the 25 September Artists 4 Safe AI letter; and a press-and-op-ed track that ran through Encode's own channels and partner platforms, including a co-authored op-ed by Revanur and Grey's Anatomy actor Jason George making the public case for signature. The campaign's tactical posture — youth-led legitimacy paired with senior-researcher validators paired with Hollywood public-letter mobilization paired with Sacramento inside work — is the template that subsequent Encode-coordinated US state-AI-policy campaigns have been built on.

Aftermath: the working group, the California Report on Frontier AI Policy, and SB 53 (October 2024 – September 2025)

The Joint California Policy Working Group on AI Frontier Models — convened by Newsom in the weeks after the veto and led by Li, Cuéllar, and Chayes — released its final 53-page California Report on Frontier AI Policy on 17 June 2025. The report recommended third-party evaluation of frontier models, incident-reporting infrastructure, whistleblower protections, and risk disclosure to downstream developers — several of which had been load-bearing provisions of SB 1047. Wiener and Encode Justice carried those recommendations into a narrower successor bill, SB 53, the Transparency in Frontier Artificial Intelligence Act. Newsom signed SB 53 into law on 29 September 2025 — exactly one year after the SB 1047 veto — making it the first US statute focused squarely on frontier-AI safety, with a $500 million-revenue scoping threshold covering roughly five to eight large developers (OpenAI, Anthropic, Google DeepMind, Meta, Microsoft), a 1 January 2026 effective date, and $1 million-per-violation civil penalties.

Significance for the broader AI-good movement

For the wider make-AI-good movement this corpus is mapping, the SB 1047 campaign matters on four connected counts. First, it is the first sustained US-state legislative campaign on the public record in which a youth-led civil-society organisation took formal bill co-sponsorship of a frontier-AI safety statute — moving Encode Justice from validator and amicus work into the inside game of state legislation. Second, the campaign assembled the broadest civil-society coalition any US-state frontier-AI safety bill had then put on the public record — combining youth advocacy with senior AI-researcher endorsements (Hinton, Bengio), AI-lab employee solidarity (the 113-plus signatories of the September letter), an entertainment-industry union (SAG-AFTRA), and a 125-plus Hollywood public-figure letter (Artists 4 Safe AI). Third, although Newsom's veto stopped the immediate signing, the bill's substantive provisions — third-party evaluation, incident reporting, whistleblower protection, risk disclosure — were carried forward into the post-veto Joint California Policy Working Group's final report and then into Wiener's SB 53, which Newsom signed on 29 September 2025, making the SB 1047 campaign legible in retrospect as a multi-year arc rather than a 2024 defeat. Fourth, the working template the campaign established — youth-led co-sponsorship plus Sacramento inside work plus validator endorsements plus a Hollywood public-letter track — has become a reference point for subsequent state-level AI-safety organising across the corpus, and is now part of the inheritance newer state-AI-policy campaigns explicitly cite.