MIRI Newsletter #126

2026-06-30 · Source: Machine Intelligence Research Institute · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, long

Summary

MIRI Newsletter #126 announces the launch of "AI StopWatch," a new daily channel for AI news and analysis, aiming to broaden engagement on AI issues beyond extinction risk. The newsletter also highlights "The AI Doc," a film MIRI supports for raising public and policymaker awareness of AI risks, now streaming on Peacock and VOD after selling approximately 80,000 tickets in its initial 10-day theatrical run. MIRI's Technical Governance Team released six papers for the ICML TAIGR workshop and two summarized papers on AI governance scenarios and a proposed international treaty for a global halt on frontier AI development. A key focus is a detailed summary of Eliezer Yudkowsky and Nate Soares' book, "If Anyone Builds It, Everyone Dies," which argues that superintelligent AI poses unmanageable risks, drawing parallels to space probes, nuclear reactors, and computer security systems, advocating for an immediate cessation of smarter AI development.

Key takeaway

For policymakers considering AI regulation, you should prioritize advocating for an international halt on frontier AI development. The inherent unfixability, rapid escalation potential, and complex vulnerabilities of superintelligent AI, as highlighted by MIRI's analysis, suggest that incremental safety measures are insufficient. Engage with global partners to establish a treaty banning advanced AI creation, mitigating catastrophic risks before they become unmanageable.

Key insights

Superintelligent AI presents unmanageable risks akin to space probes, nuclear disasters, and computer security vulnerabilities, demanding a global development halt.

Principles

• AI systems are grown, not crafted, making them opaque.
• Intelligence has narrow safety margins, risking rapid escalation.
• Complex systems are inherently insecure and exploitable.

Method

Yudkowsky and Soares propose a global halt on frontier AI development, arguing it's the only governance scenario without unacceptably large catastrophic risk.

In practice

• Advocate for a global halt on frontier AI development.
• Inform policymakers about advanced AI existential risks.

Topics

• AI Existential Risk
• Frontier AI Governance
• Global AI Halt
• Superintelligence Safety
• Machine Intelligence Research Institute
• AI Policy Advocacy

Best for: Policy Maker, AI Ethicist, General Interest

Related on AIssential

• The AI Doc: Your Questions Answered — AI's rapid, opaque development necessitates an international ban to prevent unsteerable superintelligence.
• AI Keeps Tomato Plant Alive — Uncontrolled superintelligence poses an existential risk, as current AI development prioritizes capability scaling over inherent safety engineering.
• My AI Opinions — The future of AI involves complex, probabilistic timelines for AGI, superintelligence, and critical safety and geopolitical challenges.
• Summary: AI Governance to Avoid Extinction — Only a global "off switch" and halt to frontier AI development can prevent catastrophic risks.
• Anthropic’s Dario Amodei wants governments to have the power to block ‘dangerous’ AI systems — Dario Amodei advocates for mandatory government oversight and blocking power over frontier AI systems based on compute thresholds and risk assessments.
• Prominent Scientists, Faith Leaders, Policymakers and Artists Call for a Prohibition on Superintelligence, as Poll Shows Americans Don’t Want It — A broad coalition calls for prohibiting superintelligence development until safety, control, and public consensus are achieved.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Intelligence Research Institute.