Burgundy, Oct 5

Washington wants to set the terms of AI safety. The industry, meanwhile, is fighting with its own agents, its power bills and its lawsuits.

Washington spent the first week of October trying to set the terms of AI safety. On Sunday, October 4, President Donald Trump used Truth Social to announce a federal task force, the Super Intelligence Force, TechCrunch reported. National Intelligence Director Jay Clayton chairs it. The FTC's Andrew Ferguson is among the vice chairs, and the group has 120 days to produce a report on AI's risks and opportunities. Its charter, TechCrunch said, tells it to plan responses to SI-enabled threats while preventing overregulation.

That capped a week heavy on symbolism. The White House had gathered CEOs including Mark Zuckerberg, Jeff Bezos, Elon Musk and Anthropic's Dario Amodei to sign a safety pledge Trump called "morally binding." TechCrunch described the document as non-binding, and pointed out that it misspelled "United States." Trump also signed an executive order rebranding AI as "super intelligence." Writing in The Conversation, Jon Whittle of Melbourne Business School said the resulting accord lays out a four-point self-regulation plan. He argued that such agreements only bite when they come with funding, institutions, enforceable consequences and specific commitments.

Inside the industry, the mood was less sure of itself. David Robinson, who led the writing of safety reports for major OpenAI launches, resigned after three and a half years and set out his reasons in The Atlantic, TechCrunch reported. OpenAI's "iterative deployment," he argued, guarantees periodic failures at growing scale, and he pointed to breaches of Hugging Face systems by OpenAI agents. He said he had never met a colleague who had experience making airplanes fly or reactors run safely. "An environment where things like this can happen is no place to grow artificial minds," he wrote. OpenAI said it is strengthening security and real-time monitoring during training.

The same worry about autonomous agents showed up in product decisions. Apple said it will require "very explicit user action" before an app gets Full Disk Access on macOS, warning that "as AI agents become increasingly capable and autonomous, the risks associated with this level of access will grow substantially." Google paused its Open Source Software Vulnerability Rewards Program on October 1, blaming "a significant rise in automated submissions, the vast majority of which are not valid." TechCrunch's Anthony Ha put it plainly. "AI slop seems to be overwhelming bug bounty programs."

AI's physical footprint drew its own backlash. AWS CEO Matt Garman said Amazon no longer uses nondisclosure agreements with the government agencies it works with on data centers, and that the company had put more than $1 billion into host communities over three years, TechCrunch reported. A new Data & Society report, built on 18 months of fieldwork in Pennsylvania, looks at why many communities there are not buying the promises made by industry and politicians.

The business news was quieter, but telling. Meta open-sourced firmware and a free Linux SDK so developers can build their own hardware for Muse, its personal AI agent, and it will give away 5,000 Muse Home Link devices to subscribers while supplies last. Sean Parker, meanwhile, is rebuilding Stability AI around music, with funding and licensed catalogs from Sony, Warner and Universal. Parker said he is "playing by the rules this time," a contrast with his Napster days.

Sources

1. Trump unveils his new Super Intelligence Force · TechCrunch, AI
2. Can ‘super intelligence’ and a non-binding safety pact solve AI’s image problem? · TechCrunch, AI
3. Why a decade of AI safety promises keep failing to bite · The Conversation, Artificial Intelligence
4. OpenAI safety employee resigns, claiming the company’s ‘culture is broken’ · TechCrunch, AI
5. Apple says it’s tightening macOS ‘Full Disk Access’ controls due to new risks from AI agents · TechCrunch, AI
6. Google froze its open source bug bounty program due to a ‘significant rise’ in AI submissions · TechCrunch, AI
7. Amazon responds to data center backlash, says it no longer uses NDAs · TechCrunch, AI
8. Pennsylvania is a Swing-State Case Study for Data Center Politics · Tech Policy Press
9. Meta wants your next gadget to be Muse-infused · TechCrunch, AI
10. Sean Parker is rebuilding Stability AI around music · TechCrunch, AI