2025 fundraiser
Forethought is a non-profit that researches how to navigate the transition to a world with superintelligent AI systems.3
We are looking to raise $3.9m to do follow-on work to mitigate AI-enabled coups, investigate new areas like digital rights and space governance, and build the field of people working on these issues.
We are currently in the middle of a hiring round, and have a strong hiring pool. While we’re still evaluating candidates, we think that funding might mean that we will be able to hire or support fewer researchers than we think is optimal, so funding could be unusually helpful at the moment.
Funding request
Forethought currently has funding until mid-December 2026 on our baseline budget. We need additional funding to allow us to fulfil our expansion budget, and extend our runway to June 2027.1 For details of our budget, see below.
Our funding ask is for $3.9m2 to fulfil our expansion budget to June 2027. A more basic ask would be for $2.4m to fulfil only our baseline budget to June 2027.
Marginal funding allows us to hire additional expected researchers, including senior researchers, which will significantly accelerate our research. It will also allow us to increase the budget available for our impact function, to conduct activities such as run fellowships and events to build up the research and policy fields around our ideas, which will make sure that our ideas lead to real changes in AI development.
If you have any questions, please feel free to contact us.
Other funders
Forethought has been evaluated and endorsed by both Longview and by Coefficient Giving (formerly Open Philanthropy), who are recommending us to donors in their network.
Why Forethought?
See our Donate page for testimonials, details on the case for Forethought, our theory of change and achievements so far.
Our plans for 2026 and onwards
Research
Our research topics in 2026 are likely to include:
- How to improve AI character
- What AI model specs should look like, including questions like how AI systems should respond to power-seeking requests, how they should approach ethical reasoning, and whether they should be more guided by instructions or by virtues.
- This has included some advising of companies on model spec, and may grow into broader work on AI character that doesn’t flow through the model spec in particular.
- Follow-up work on AI-enabled coups
- Further research on threat modelling and concrete recommendations for AI companies and governments
- Further work with AI companies to ensure they implement safeguards against power-grab risk
- Building on our work on Better Futures to get clearer sense of next steps, including questions like how important moral diversity is, whether liberal democracy is sufficient for reaching a near-best future, and whether and how to hand off decisions about the future to future generations.
- Deals with AIs
- Exploring in what circumstances deals with AIs can be used to reduce the risk of misaligned AI takeover
- Space governance and strategy
- Working with Toby Ord and an external space governance experts to scope out the potential for strategy and policy work in space governance
Impact
While we’re pleased with the impact of our research so far, we want to more systematically ensure that our research leads to action (and quickly). Our plans for 2026 include:
- Continue the follow-up work to AI-enabled coups
- Facilitating technical research on detecting "secret loyalties" in AI systems
- Working with companies to integrate system integrity safeguards into AI security frameworks
- Handing off military policy work on command structure vulnerabilities for autonomous systems to better-positioned organizations
- Coordinating other efforts in this area
- Build the field of people thinking about how to navigate rapid AI progress wisely
- Opportunistically explore new areas where we can find excellent people to work on poorly-scoped but important problems. This could include:
- Space governance
- Deals with AIs
Hiring
Forethought is currently evaluating applicants for (Senior) Research Fellow positions, which we advertised for in October. (Applications are now closed.) We have been impressed by the quality of applications. We expect to hire at least 3 researchers, though it may be more, depending in part on funding.
We aim to build a strong, diverse intellectual community in our field, and think that we could valuably collaborate with and support several others from our current hiring round and more broadly.
Marginal funding helps Forethought hire additional researchers, and determines the extent to which we’re able to support our broader intellectual field, for example with a visiting fellows program, grants or other support.
2026 budget
Image
Notes:
- For the calculation of our funding gap until June 2027, we project forward a constant spending rate from the end of 2026.
How to donate
- You can donate via the button below:
- Donations via this button are processed via Every.org:
- Such donations are tax-deductible for donors in the US.
- Donors based in the US can donate via credit card, bank transfer, PayPal, Venmo, cryptocurrency, stocks, or from a donor-advised fund (DAF).
- Donors based outside the US can donate via credit card, PayPal or crypto.
- If you’d like to make a (non-tax-deductible) donation via bank transfer from outside the US, please contact us and we can facilitate that.
- If you’d like to make a donation above $20,000 via a method not listed above or needing tax-deductibility in another country, please contact us and we can look into it.
- It's plausible that we'll have a route to receive UK donations with Gift Aid in late 2025 / early 2026. If you'd like to be notified if/when this becomes possible, please fill in this form.