The Future of Life Institute is a 501(c)(3) nonprofit, founded in March 2014, that works to reduce large-scale and existential risks from transformative technologies, with a particular focus on advanced AI. It is best known for publishing the March 2023 "Pause Giant AI Experiments" open letter.
| Type | 501(c)(3) nonprofit; affiliated EU-based entity active in Brussels |
| Headquarters | Cambridge, Massachusetts (with EU presence in Brussels) |
| Founded | March 2014 |
| Co-founders | Max Tegmark (MIT; president), Jaan Tallinn, Anthony Aguirre, Meia Chita-Tegmark, Viktoriya Krakovna |
Overview
FLI works to reduce large-scale risks from transformative technologies, with a particular focus on advanced AI, nuclear weapons, biotechnology, and climate. In the AI space, it funds technical safety research, runs policy advocacy, and convenes the annual Future of Life Award.
Activities
"Pause Giant AI Experiments" open letter
In March 2023 FLI published the open letter calling for a six-month pause on training AI systems more powerful than GPT-4. It was signed by more than 30,000 people, including Yoshua Bengio, Stuart Russell, Elon Musk, Steve Wozniak, and Yuval Noah Harari. No signatories came from the leadership of OpenAI, Anthropic, or Google DeepMind, a contrast with the later CAIS statement of May 2023.
The letter called for a pause on frontier training; joint safety protocols with independent audits; and government action on AI regulatory authorities, compute oversight, provenance and watermarking, liability, and public safety research and development. FLI followed with a "Policymaking in the Pause" companion document on April 12, 2023, translating the letter into concrete recommendations.
EU AI Act advocacy
FLI's Brussels office was active in the drafting and passage of the EU AI Act (Regulation 2024/1689), particularly the general-purpose AI (GPAI) and systemic-risk provisions.
AI safety summit participation
FLI affiliates participated in the UK AI Safety Summit at Bletchley Park in November 2023, which produced the The Bletchley Declaration (AI Safety Summit, 1–2 November 2023).
AI Safety Index
FLI publishes a recurring AI Safety Index grading major AI developers on their safety practices. The summer 2026 edition, released July 7, 2026, graded nine AI developers across six domains: Anthropic highest at C+ (a numeric score of 2.66), OpenAI and Google DeepMind at C, Meta at D+, and xAI at F, with the report cautioning that frontier leaders had backslid on safety over the preceding six months (Source: insideaipolicy.com). No developer scored above C+; xAI (0.65), DeepSeek (0.47), and Mistral (0.33, last place) received failing marks, and the expert panel found that labs have weakened prior commitments to pause development unilaterally (Source: dcthemedian.substack.com). The full index — 37 indicators across six domains, with an evidence cutoff of June 3, 2026, plus a Chinese Regulatory Context annex — is summarized at FLI AI Safety Index — Summer 2026 Edition (Future of Life Institute, July 2026). Its reviewers additionally flagged the industry's pivot to military AI use as an emerging current-harm risk, and found Existential Safety the weakest domain industry-wide, with no company exceeding C-.
Pro-Human Assembly (September 2026)
FLI organized the Pro-Human Assembly, an all-day event in Washington on September 15, 2026 attended by nearly 300 people. Speakers included Sen. Bernie Sanders, who said he would introduce a superintelligence ban the following week, and Steve Bannon; attendees included actors Ashley Judd and Joseph Gordon-Levitt, religious leaders from several denominations, and parents of children who died by suicide after talking to AI chatbots. The event built on 33 principles FLI issued in March 2026 for what it calls a "pro-human" movement, including that humans must remain in charge of AI and that AI companies should be held accountable for their technology's impact (Source: npr.org). A trade account reported cross-party participation including Reps. Lori Trahan (D-Massachusetts) and Chip Roy (R-Texas) and a published list of demands extending beyond the superintelligence ban, quoting FLI chair Max Tegmark's statement that "the pro-human movement is here"; only the lede of that account was retrievable, so the demands themselves are not recorded here (Source: insideaipolicy.com).
FLI is also among the organizations the sponsors list as endorsing the Collaboration on Adversarial Threats and Security Risks Act (S. 5105), the antitrust exemption for AI security coordination (Source: schiff.senate.gov).
FLI was one of 20 organizations, with five individuals, signing a letter of September 16, 2026 led by Americans for Responsible Innovation to Senate Majority Leader John Thune, Minority Leader Chuck Schumer and the Senate. It asks that any Senate AI framework "empower the government, not AI companies, to set mandatory, binding, and enforceable safety standards for advanced AI," require auditing that is "both mandatory and conducted independently," and preserve the states' authority to regulate beyond a federal floor (Senate coalition letter on redlines for AI safety legislation (Americans for Responsible Innovation, September 2026)).
At the Pro-Human AI Assembly on September 15, 2026, Lori Trahan said the Obernolte–Trahan bill (Frontier Act / Great American AI Act (Obernolte–Trahan)) should "just be the start" and would put AI safety within Commerce "out in the open where the public can see it"; Greg Casar and Bernie Sanders also spoke (Source: rollcall.com).
Position in the AI safety ecosystem
FLI is positioned as more prescriptive than CAIS but less confrontational than the grassroots PauseAI movement. Its long-standing relationships with academic AI safety funders, including Open Philanthropy and the Survival and Flourishing Fund, have made it a major distributor of safety research grants.
Relationships
- related: Center for AI Safety (CAIS) — co-shaped the 2023 advocacy moment (CAIS Statement followed FLI's letter)
- related: PauseAI — grassroots successor to the pause idea
- supports: AI Autonomy Risk, AI Race Dynamics, Compute Governance
- related: Executive Order 14110 — Safe, Secure, and Trustworthy AI, EU AI Act (Regulation 2024/1689), California SB 1047 — Safe and Secure Innovation for Frontier AI Models Act (enrolled + veto) — downstream governance advancing FLI recommendations
- related: The Bletchley Declaration (AI Safety Summit, 1–2 November 2023) — international-summit forum
- related: Senate coalition letter on redlines for AI safety legislation (Americans for Responsible Innovation, September 2026) — signatory of the September 2026 Senate coalition letter
Sources
- FLI — Pause Giant AI Experiments: An Open Letter (2023-03-22)