Future of Life Institute (FLI)

What it does

Advocacy, grantmaking and policy across AI, biotechnology and nuclear weapons. Outputs include the AI Safety Index (latest edition Summer 2026, published July 2026, grading nine companies — Anthropic C+, OpenAI C, Google DeepMind C, Meta D+, Z.ai D-, Alibaba Cloud D-, xAI F, DeepSeek F, Mistral F — reviewed by a seven-member panel of David Krueger, Sharon Li, Tegan Maharaj, Sneha Revanur, Stuart Russell, Robert Trager and Yi Zeng), the 2023 pause letter, EU AI Act advocacy, and substantial grant programmes.

What they ask for — in their own terms

A moratorium on developing artificial superintelligence for at least 15 years, plus licensing, auditing and enforceable safety standards before scaling — and a participatory global process rather than decisions made by a handful of CEOs.

What they do

Open letters and petitions (Pause Giant AI Experiments, 31,810 signatories; Asilomar AI Principles), a contact-your-legislator tool, EU AI Act implementation work, grantmaking, the Digital Media Accelerator, and a newsletter with about 70,000 subscribers. It is a registered lobbying entity in both the US and EU.

Honest assessment

Unusually responsive for its size, and the only major org here that publishes an explicit response-time commitment. Well-resourced, so it can act on what it receives. The Index gives it leverage over company behaviour that few others have.

What a newcomer can do

Use the contact-your-legislator tool, sign the open letters, join the action alerts list, or apply to the Digital Media Accelerator if you make creative content.

Worth knowing

FLI does not have a separately branded advocacy arm — its policy and lobbying work runs through the Institute itself plus EU-registered activity. Its ask is a time-limited moratorium on superintelligence, which is weaker than ControlAI's or StopAI's permanent ban and stronger than Encode's transparency agenda. It is also a funder of other groups on this list, so 'independent' endorsements between these organisations are not always independent.

Concerns it covers

Extreme large-scale and existential risk, with loss of control and uncontrolled capability escalation as the central AI concern. Also carries structural concerns (power concentration, autonomous weapons). Not a near-term-harms organisation, though the Index rates some near-term practices.

Government channel

Strong — EU Transparency Register presence and sustained EU AI Act engagement, plus US state and federal advocacy and UN input.

Tips

no

Where they work

United States and European Union (Brussels lobbying presence).

Size

Mid-sized and well-funded by the standards of this field; a large grantmaker as well as an advocate.

Spending

Registers as a federal lobbying client in the US (OpenSecrets ID D000087562) and appears on the EU transparency register; specific figures not retrieved from its own pages.