AIToday
AI Safety & AlignmentAI Business & IndustryMIT Technology Review AIPublished: Aug 26, 2026, 19:00 JST2 min read

Gates: AI dangers have crossed thresholds

Gates: AI dangers have crossed thresholds

Key takeaway

  • Bill Gates says AI has crossed safety thresholds in bio, cyber, jobs, and control. He finds society's lack of concern shocking.

  • He urges monitoring of models that can make new molecules.

  • Bioterrorism risk, he says, is about 50 times more likely than a natural pandemic.

3 Key Points

  1. What happened

    Bill Gates has published a new essay and given interviews warning that AI's capabilities have passed several danger thresholds, including bioterrorism, cyberattacks, job-market disruption, and psychosocial dependence, and hints at loss of control.

  2. Why it matters

    Gates says society outside the industry is not paying enough attention, and he views bioterrorism risk as about 50 times more scary and likely than a natural pandemic. He is calling for monitoring of models that can make novel molecules and for new ideas like human-reserved jobs and taxes on robots and tokens.

  3. What to watch

    Gates predicts that for a substantial part of white-collar work, AI will be cheaper and better, including almost every entry-level job, and he expects this to happen soon, as capability and reliability issues are being solved.

Ask the AI about this article →

Context & Analysis

Bill Gates's warnings come as he observes that AI capabilities have advanced past thresholds that were previously seen as future checkpoints. He emphasizes that while the industry is aware, there is a lack of concern and discussion outside it, which he finds stunning. His comparison of bioterrorism risk to natural pandemic risk—calling it about 50 times more scary and likely—underscores his urgency.

Gates also counters the argument that past technological revolutions did not reduce net jobs, saying this time is different because AI can replace human cognition for a high percentage of jobs across every industry at modest cost and with lower error rates. He points to signs like some companies hiring fewer entry-level workers as early indicators. He proposes forward-looking ideas such as human-reserved jobs and taxes on robots and tokens to address the transition.

Despite the turbulence he predicts, Gates remains optimistic about AI's benefits in agriculture, health care, education, and navigating bureaucracy. He mentions the Gates Foundation's work in vaccines and drugs and a spin-off called NextLadder to help low-income families with benefits and training programs. His message balances concern with a belief that eventually society reaches abundance, but only after significant disruption.

FAQ

What does Bill Gates say about AI and jobs?
Gates says AI is now cheaper and better for a high percentage of white-collar jobs, including almost every entry-level job, and that the past is misleading because this time AI can replace human cognition at modest cost.
What is Gates's proposal for monitoring AI models?
Gates says any model that can make novel molecules should be monitored and not be copyable into a place where monitoring logic is removed. He suggests the US should declare such models subject to monitoring and approach China for agreement.
MIT Technology Review AIRead Original Article

Get the latest AI Safety & Alignment news every morning

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytime

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleApple takes AI compute out of data centers with new Macs