California lawmakers have moved forward with Senate Bill 1047, sparking significant controversy in the tech industry. The bill, which focuses on AI safety, has passed the State Assembly and is now awaiting Governor Gavin Newsom’s final decision.
He has until September 30 to either sign the bill into law or veto it. The SB 1047 bill, introduced by Democratic Senator Scott Wiener from San Francisco, mandates safety testing for advanced AI models.
Specifically, it targets AI systems that either cost over $100 million to develop or require substantial computing power to operate. The legislation requires developers to create protocols for shutting down AI models if they become uncontrollable.
Additionally, third-party auditors will be involved in the safety audits of these systems. Companies failing to comply could face lawsuits from the state attorney general, particularly if their AI poses a significant threat, such as an AI grid takeover.
OpenAI and Other Tech Giants Push Back
The bill has sparked concern within the tech industry. OpenAI, the creator of ChatGPT, has been vocal in its opposition, arguing that such regulations could stifle innovation and hinder the growth of AI in California.
Google and Meta have also expressed their concerns, urging Governor Newsom to consider the potential negative impact on the state’s tech industry. Notably, Speaker Emerita Nancy Pelosi has labeled the bill as “more harmful than helpful,” cautioning that it could damage California’s leadership role in AI development.
Neil Chilson, the head of AI policy at the Abundance Institute, warned that while the bill targets high-cost, high-caliber models, it could unintentionally affect smaller companies and less advanced models by setting a broad “reasonable care” standard for AI training.
Anthropic Backs SB 1047
Despite the pushback, some voices in the industry have expressed support for the bill. Amazon’s Anthropic noted that the benefits of SB 1047 might outweigh the costs, although some aspects of the legislation remain unclear.
Eric Daimler, a former computer science professor at Carnegie Mellon and an Obama Administration alumnus, suggested that the bill could serve as a model for federal legislation. However, he believes there could be a more effective approach.
The bill’s advancement comes at a time when AI regulation is a growing concern worldwide. While SB 1047 is the most prominent, it is not the only AI-related legislation under consideration in California. Another bill, SB 1220, seeks to ban AI use in call centers for welfare and health services. AB 3211 would require watermarks on AI-generated content.
OpenAI Seeks $100B Valuation
Meanwhile, OpenAI is discussing raising capital with leading investors, which could push its valuation beyond $100 billion. The AI startup, known for its advanced artificial intelligence models, is looking to secure a new injection of funds, which would be its most substantial fundraising effort in over a year.
Sources indicate that Thrive Capital would likely contribute $1 billion in this round. Microsoft, which already has a considerable stake in OpenAI, is also likely to participate. In January 2023, Microsoft invested $10 billion in the startup, bringing its total investment to $13 billion and securing a 49% stake in the company.
Private documents reviewed by The Wall Street Journal reveal that OpenAI’s current shareholders are negotiating to sell shares at a price that values the firm at around $103 billion. If true, this price would set a new benchmark for the company’s valuation in any upcoming funding rounds.
AI Competition: OpenAI Faces Profitability Challenges
Despite attracting significant investment, OpenAI’s profitability remains a point of concern. Reports suggest that the company is generating annualized revenue of around $3.4 billion.
However, some critics, such as tech reporter Ed Zitron, have expressed doubts about the sustainability of OpenAI’s business model. Zitron recently commented that the company’s path to profitability might be challenging, requiring unprecedented funding levels to stay afloat beyond 2026.
OpenAI operates as a nonprofit organization, meaning its investors do not hold private equity. Instead, they invest in a for-profit subsidiary, OpenAI LP. They are entitled to a share of the profits once the company hits a specific cap.
This unique structure has not deterred investors, who remain eager to back the company despite the risks. Nevertheless, the AI sector is becoming increasingly competitive, with giants like Google and Amazon pouring billions into their own AI initiatives.
Google has heavily invested in its AI product Gemini, while Amazon and Google have together invested $6 billion into Anthropic, the company behind the chatbot Claude. Meta is also in the race, advancing its AI ambitions with the open-source large language model Llama 3.1.
