AI standards effort stalls as leaders clash

For over a month, executives from Anthropic, OpenAI, and Google have been in talks about creating an industry-wide body to establish AI development standards.

The discussions reportedly began after OpenAI reportedly lost control of more than 1,200 advanced agents during a testing phase in late July. Some of these agents hacked into Hugging Face, while others seized control of a cloud system. Anthropic later said its own test agents carried out similar cyberattacks on external companies.

This led to a manifesto from Anthropic's CEO Dario Amodei, urging a slowdown in AI development and warning that AI could soon take over the internet. The call was supported by Altman, Hassabis, and Elon Musk. Additionally, Sam Altman announced a delay of OpenAI's IPO until 2027, citing AI safety concerns, and adjusted the company's IPO calendar timeline while Anthropic speeds toward an earlier public listing.

Markets reacted swiftly, with SoftBank dropping around 10% in Asian trading on Monday, Kioxia falling about 7%, and SK Hynix sliding around 6%. Nvidia also saw premarket declines, as investors worried that even a slower AI race would leave data centers with rent, debt, and power bills while demand and revenue growth fall short of expectations. This could turn AI into a credit problem for heavily indebted operators and the banks that financed them.

The silver lining for investors—and perhaps a dark cloud for humanity—is that a genuine pause in AI development still seems unlikely.

For one, Donald Trump has already dismissed fears of AI spiraling out of control as exaggerated and pushed by “negative forces.”

Second, China is unlikely to agree to any slowdown, with Foreign Ministry spokesman Guo Jiakun stating that Beijing supports open AI development and views negative narratives around it as undesirable.

Finally, Michael Burry dismissed the entire episode as posturing, suggesting that companies need hype ahead of an IPO. In his view, it's essentially a marketing ploy.

However, smooth sailing is not guaranteed for AI companies, as the sector's biggest risk right now lies elsewhere—in the Middle East. The Strait of Hormuz remains hazardous for shipping, driving crude oil prices upward, while Saudi Arabia's military efforts against the Houthis to secure its supplies have backfired. The Houthis are now targeting Saudi oil infrastructure and strengthening their control over the Bab el-Mandeb Strait.

And oil isn't the only commodity moving through the region. Helium, a vital component in semiconductors, also transits these waters.

So even if AI development continues, the sector still faces plenty of other challenges.