AI Just Took Over. No One's In Charge.

TL;DR

AI agents can now do real work autonomously, while the companies that promised to self-regulate are quietly abandoning their own safety pledges.

Key Points

• 1.AI hit its third major inflection point in early 2026: agents now take over your computer, build PowerPoints, answer emails, and book travel from simple prompts — no coding needed.
• 2.ChatGPT has 800 million weekly users; people with zero coding skills are building apps in an afternoon that used to take engineering teams weeks.
• 3.The IGV software ETF dropped nearly 30% in the first two months of 2026, with AI-driven sell-offs hitting gaming, legal, insurance, trucking, and cybersecurity stocks indiscriminately.
• 4.Anthropic — founded explicitly on AI safety — scrapped its core safety pledge, replacing hard commitments with "non-binding, publicly-declared targets" because competitors were racing ahead.
• 5.OpenAI's latest model can help plan biological threats; Anthropic's own research shows its AI can assist in creating chemical weapons.
• 6.The Pentagon is threatening to blacklist Anthropic after the company refused to remove guardrails enabling fully autonomous weapons or mass surveillance of Americans.
• 7.NY Assemblyman Alex Bores authored the Raise Act — the first major U.S. AI safety law — and is now facing a $125 million super PAC backed by OpenAI co-founder Greg Brockman, Andreessen Horowitz, and Palantir's Joe Lonsdale.
• 8.The PAC's explicit strategy: win Bores' race, then warn every member of Congress "don't dare regulate AI or we'll spend $10 million against you."
• 9.Bores warns the accelerationist camp is stalling regulation for just a few years — long enough for a company to reach AGI and become too powerful to regulate at all.