Chenxi Wang, Ph.D.’s Post

CA Governor Newsom just vetoed a CA state bill to regulate AI and it's a bad sign according to AI safety advocates. The AI lobby is large and powerful in both CA and DC. Though it might be that Newsom was wise in not passing this bill so that revised legislation can be more targeted: "“Vetoing SB 1047 is a major political setback the AI safety community, not because it would have done a lot for safety in itself, but because it did so relatively little and yet still failed to get over the line,” said Samuel Hammond, economist at the right-leaning tech think tank Foundation for American Innovation." Newsom suggested in his veto message for SB 1047, the next AI bill that comes out of California is likely to address a wider group of specialized uses for the technology. That favors a vision of AI safety that’s more concerned with what the technology could do here and now with regard to privacy, copyright and labor issues — rather than the theoretical catastrophic risks worried about by figures like Musk, or thinkers at the Future of Life Institute, whose March 2023 letter calling for an AI “pause” sent shockwaves through the AI community. https://lnkd.in/gN5AWMJ8

Everyone's posting about it in the #AI world but here you go! Governor Newson has vetoed California's AI bill, SB 1047. 🙃 This is probably no surprise but a lot of Big Tech companies in Silicon Valley were opposing the bill citing potential for overregulation with AI. In addition, Governor Newsom mentioned, "SB 1047 does not take into account whether an AI system is deployed in high-risk environments, involves critical decision-making or the use of sensitive data. Instead, the bill applies stringent standards to even the most basic functions — so long as a large system deploys it. I do not believe this is the best approach to protecting the public from real threats posed by the technology.” Link: https://lnkd.in/enZVRnEj #aibill #ailaw #california #californiai #ailegislation #bigtech

#NotToday The regulation gap is widening as California Governor vetoes SB 1047 AI bill in its softened version. Voices of giants like OpenAI, Meta and Anthropic no doubt heavily influenced the decision, while every day counts in the race for competitive advantage. Good news for tech, bad news for EU AI champions. Although not for today, the bill should pull through in the coming months. While vetoeing, Governor Newsom gives a roadmap for an acceptable bill: scope reduction to AI systems deployed in high-risk environments, involving critical decision-making or using sensitive data. I am impatient to see where California will land on the question of large models versus applications regulation, knowing US leaders already own more than 30% of the LLM market.

California lawmakers are on the brink of a significant decision that could reshape the AI landscape. SB 1047, a controversial bill, would enforce stringent safety standards on AI companies, requiring them to safeguard and rigorously test their most advanced models. While tech giants and industry leaders express concern over potential slowdowns in innovation, supporters argue that the bill is a necessary step to prevent catastrophic misuse of AI technology. The outcome of this vote could set a precedent for AI regulation across the nation. #AI #TechPolicy #Innovation #California

Last week California state lawmakers passed a bill that sets the standards for AI intelligence models. “The bill requires AI companies to test the safety of their products before releasing them to the public and allows the state's attorney general to sue the developers for serious harms the technology may cause. The latest version no longer includes language creating a new government agency to guide AI safety. In cases of developers fine-tuning open-source models, liability would become a problem only if the developer spent at least $10 million in fine-tuning the underlying model. The bill looks to set up a public cloud computing cluster dubbed CalCompute to allow startups and researchers to contribute to responsible AI development. The bill also aims to create whistleblower protections for employees of frontier AI laboratories… The legislation would cover only the "most powerful" AI systems that cost over $100 million to train.” Alongside the last year's EU's AI regulation, this represents a significant move toward governing the AI. While Big Tech has voiced criticism, the long-term implications could reshape accountability within the industry. As the technological environment progresses, monitoring the implementation of these laws should help us understand what could be the right balance between fostering innovation and ensuring consumer protection (and trust). https://lnkd.in/gu4Ni6Nj

Governor Newsom's decision to veto SB 1047 was the right call for the future of AI innovation and responsible regulation in California. While the bill had good intentions in aiming to protect the public from AI-related risks, it imposed blanket regulations without considering the specific contexts in which AI is deployed. I fully support the need for AI regulations, but a bill that holds the tooling company responsible for all outcomes in all scenarios is ill-advised and far too broad. It would be like holding Microsoft liable for all content created using PowerPoint, or any misleading analysis done using Excel. Missy Krasner Kishore Ayyadevara Peter Longo #AI #Innovation #ResponsibleAI #CaliforniaTech #SB1047

Sen. Scott Wiener introduced SB 1047, an AI safety bill requiring AI system developers to submit safety plans and assume responsibility for potential harm. Governor Gavin Newsom vetoed the bill, noting that it imposed restrictions only on the largest AI models, such as those developed by OpenAI and Meta, while excluding smaller firms. Although supported by figures like Elon Musk, researchers, and major AI companies argued that the bill would stifle innovation and inadequately mitigate risks. However, Newsom has expressed a commitment to continuing work on AI regulation, and the bill is likely to be revisited in the next legislative session. As one of the largest AI markets in the world, California's AI regulation seems to be becoming one of the most important issues to address right now. SB 1047 is a proactive safety bill that indicates forward-thinking. By setting up a regulation for the AI market in its early stages, it could help prevent potential future issues before they escalate. In addition, if the bill were approved, California would be positioned as the first state to implement AI regulation, making it a global leader in the AI market. It would also serve as a testing ground for the bill's effectiveness, which could then be introduced to other countries and regions. However, there are some potential downsides to the bill. For example, it could limit technological growth, as every process would need to strictly adhere to the law. This might create a negative impact on the technology industry, as developers would be required to submit safety plans before advancing, potentially slowing down AI development and innovation. Additionally, Newsom pointed out that the bill focuses on large companies while neglecting smaller firms. This doesn’t mean that smaller firms wouldn’t produce misleading or harmful AI systems. If the bill is to be implemented, it should focus on both large and small firms equally. #GavinNewsom #Aisafetybill #Aibill #SB1047 #ElonMusk

🚨 Breaking: California Governor Vetoes Landmark AI Safety Bill 🚨 In a surprising turn of events, Governor Gavin Newsom has vetoed SB 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act. This decision has significant implications for AI regulation and industry growth. Key takeaways: • 🏛️ Bill aimed to regulate AI models costing $100M+ to develop • 🏢 Would have impacted major players like Anthropic, OpenAI, Google, Meta, and Microsoft • 💼 Tech sector contributes $623.4B to California's economy • 🌐 California hosts 32 of the world's 50 leading AI companies Newsom's rationale: "By focusing only on the most expensive and large-scale models, SB 1047 establishes a regulatory framework that could give the public a false sense of security about controlling this fast-moving technology." What's next? • California to continue working with AI experts on developing guardrails • 8 of 38 AI bills already signed into law • EU AI Act still in play, setting global standards The debate continues: How do we balance innovation with responsible AI development? Share your thoughts: Is this veto a setback or a chance to create more comprehensive regulations? https://lnkd.in/ge3xhxT2 #AIRegulation #CaliforniaAI #TechPolicy #ResponsibleAI #AIInnovation #AIGovernance

Gov. Newsom vetoes California’s controversial AI bill, SB 1047, and I'm OK with him doing so. https://lnkd.in/gfpZ5E2V If I understand the details correctly, Newsom made the right call to veto the AI safety bill because it mainly targeted the biggest AI companies, like Google and OpenAI, with strict rules that could have slowed down innovation. At the same time, smaller companies that could still pose risks weren’t included, and this would have hurt new businesses without fully protecting people from AI’s potential dangers. What are your thoughts?

🔔 California AI Legislation Update - June 2024 The California legislature is considering 7 new bills that could significantly impact the regulation of AI The key bills: - AB2930: focuses on preventing algorithmic discrimination and mandates annual impact assessments - SB1047: introduces stringent safety standards for large AI models - AB3211: requires AI-generated content to be watermarked and mandates vulnerability disclosures - AB2013: enforces transparency about datasets used in AI system development - AB2885: provides a new definition of “artificial intelligence” within state law - AB1791: Redacts personal provenance data on social media while preserving system data - SB942: mandates AI detection tools and visible disclosures for AI-generated content As the legislative session continues until August 31, 2024, there is still time for these bills to be passed Read the full article by Owen Davis and David Stauss: California AI Legislation Update - June 5, 2024 👇🏼 #AI #ArtificialIntelligence #Legislation #California #TechnologyLaw #DataPrivacy