California Senator Introduces Legislation to Protect AI Whistleblowers and Boost Responsible AI Development
SAN FRANCISCO, February 28, 2025 – Senator Scott Wiener (D-San Francisco) has introduced Senate Bill 53 (SB 53), a legislative initiative aimed at fostering the responsible development of large-scale artificial intelligence (AI) systems. The bill addresses growing concerns about AI risks while seeking to capitalize on its potential benefits.
AI is rapidly transforming various sectors, offering advancements from mental health diagnostics to wildfire response. However, leading AI researchers, including figures like Nobel laureate Geoffrey Hinton, have raised alarms about the technology’s inherent risks, advocating for immediate safeguards.
SB 53 seeks to strike a balance between advancing AI development and implementing crucial safety measures. The bill enhances protections for whistleblowers, considered vital in alerting the public to emerging AI risks. Furthermore, it initiates the establishment of CalCompute, a research cluster designed to support startups and researchers developing large-scale AI applications. This move is intended to help California maintain its global leadership in AI, particularly as states like New York invest in their own AI research initiatives.
When Governor Newsom vetoed SB 1047, Senator Wiener’s previous AI legislation, the Governor announced the state’s collaboration with a group of academics and civil society leaders to formulate AI safeguards. This working group is slated to release an interim report in the coming weeks or months, followed by a final report later this year.
“The greatest innovations happen when our brightest minds have the resources they need and the freedom to speak their minds. SB 53 supports the development of large-scale AI systems by providing low-cost compute to researchers and start-ups through CalCompute. At the same time, the bill also provides critical protections to workers who need to sound the alarm if something goes wrong in developing these highly advanced systems,” said Senator Wiener.
“We are still early in the legislative process, and this bill may evolve as the process continues. I’m closely monitoring the work of the Governor’s AI Working Group, as well as developments in the AI field for changes that warrant a legislative response. California’s leadership on AI is more critical than ever as the new federal Administration proceeds with shredding the guardrails meant to keep Americans safe from the known and foreseeable risks that advanced AI systems present.”
California currently boasts the largest concentration of leading AI companies and researchers globally. However, other regions are making significant investments in the field to gain a competitive edge. New York Governor Kathy Hochul recently announced over $90 million in new funding for the state’s AI consortium, focused on advancing responsible AI innovation.
Research clusters can be instrumental in driving innovation. The development of advanced AI models requires substantial computing resources, often inaccessible to startups and independent researchers. SB 53 aims to address this by offering low-cost computing and other resources to support responsible AI development.
“California has a history of making important public investments where it counts: from stem cell research to our stellar higher education system, we have led the way in using public dollars to foster the American entrepreneurial spirit,” said Teri Olle, Director of Economic Security for California Action. “We are proud to sponsor SB 53, a bill that would continue in this important tradition by creating the infrastructure for a public option for cloud computing needed in AI development. If we want to see AI used to promote the public good and make life better and easier for people, then we must broaden access to the computing power required to fuel innovation.”
“With CalCompute, we’re democratizing access to the computational resources that power AI innovation,” said Sunny Gandhi, Vice President of Political Affairs at Encode. “And by protecting whistleblowers, we’re ensuring that security isn’t sacrificed for speed. California can be a leader by making transformative technology both more accessible and more transparent.”
The current federal administration has reversed executive orders implemented by the previous administration establishing safeguards for advanced AI systems. The National Institute of Standards and Technology, which sets safety standards for AI, is also preparing for potential layoffs.
AI researchers continue to advocate for the implementation of guardrails to mitigate AI risks. In a significant development, Professor Yoshua Bengio, a prominent figure in AI, unveiled the world’s first International AI Safety Report, a 298-page document compiled by 100 independent AI experts from around the globe. Whistleblowers, employees of major AI labs who report on safety and security concerns, are increasingly viewed as a key source of information about the practices of large AI labs and the capabilities of large-scale AI models.
SB 53 acknowledges the need for a comprehensive strategy to foster responsible AI innovation. The bill seeks to achieve this by:
- Ensuring that employees of frontier model labs are shielded from retaliation for reporting concerns about critical risks or the dissemination of false or misleading information regarding risk management.
- Establishing a process for creating CalCompute, a public cloud-computing cluster dedicated to researching the safe and secure deployment of large-scale AI models.
SB 53 is supported by Encode, Economic Security Action CA, and Secure AI.
