Lawmakers in California have voted to advance legislation which would implement higher levels of testing and safety measures for AI companies. The goal of the legislation is to prevent AI tools from being manipulated for malicious purposes. Experts have identified possible scenarios in which AI systems could be manipulated to wipe out a state’s electric grid or to help build chemical weapons. The bill is strongly opposed by tech firms such as Meta and Google, who say that the regulations will harm developers instead of going after bad actors. The new requirements would only apply to systems costing more than $100 million in computing power to train, and no current AI models have reached this level of expense.