Leading AI safety organization drops technical research to focus exclusively on policy

The Machine Intelligence Research Institute (MIRI) is shifting its strategic focus from technical AI alignment research to public policy and communications, reflecting growing concerns about the risks of superintelligent AI.

Strategic pivot and mission evolution: After 24 years focused on technical AI safety research, MIRI has fundamentally changed its approach due to concerns about the pace of AI development and potential catastrophic risks.

• The organization now believes that without international government intervention to pause frontier AI research, catastrophic outcomes are highly probable
• MIRI’s new priority is educating policymakers and the public about AI risks rather than pursuing technical solutions
• The institute maintains some AI alignment research but has significantly scaled back these efforts since 2020-2021

Organizational developments: MIRI is expanding its team and launching several new initiatives as part of its reformed strategy.

• New projects include technical governance papers, AI risk educational resources, and an upcoming book
• The organization is actively recruiting communications specialists and technical governance researchers for 2025
• Multiple new teams have been established to support the revised mission and strategy

Financial status and funding landscape: Despite healthy reserves, MIRI faces uncertainty about future funding as it transitions to its new strategic direction.

• Current reserves of $16M provide approximately two years of operational funding
• Annual expenses have ranged from $5.4M to $7.7M between 2019-2023
• Projected spending for 2024 is $5.6M, with 2025 estimates between $6.5M and $7M
• Market performance in 2024 and lower-than-expected expenses have helped maintain reserves

Future outlook: The organization’s ability to sustain its new direction hinges on donor response to its strategic shift.

• The simpler, less technical nature of MIRI’s new mission could potentially attract broader donor support
• Increased public attention to AI risks might facilitate fundraising efforts
• However, uncertainty remains about whether previous donors will continue supporting the organization’s new focus on communications and policy

Strategic implications: MIRI’s transformation from a technical research organization to a policy and communications-focused entity represents a significant shift in how some AI safety organizations are approaching existential risk, suggesting growing skepticism about purely technical solutions to AI safety challenges.