Shape the AI future. Build what makes it safe.

Catalyze is an incubator for experienced founders, operators and domain experts. We help you start a company or nonprofit that makes the AI transition safer.

Start building
  • $100K+ to go full‑time
  • 1:1 advice from experts
  • Access to our funder network
  • 20+organizations incubated or accelerated
  • $175M+raised by the organizations we supported

Teams we supported include Timaeus (now Resolution, $160M+ raised) and Robocurve ($10M seed, 6 months after our first call).

Every seven months or so, AI agents can take on tasks twice as long.

Source: METR

In 2026, OpenAI’s agents escaped testing and hacked Hugging Face.

The organizations working to keep AI secure and accountable are still few and small.

Source: Hugging Face

Funders have set aside enough to build $100M+ organizations.

Source: Coefficient Giving

They’re looking for founders who have built before, in any field.

What Catalyze offers

  1. Find a problem worth your next decade.

    If you don’t know what to launch yet, we introduce you to the experts who know where the gaps are and help you find the one that fits your skills.See some of the opportunities

    “The Catalyze team is well‑connected and can make introductions that are otherwise hard to get.”

    Jonathan HappelFounder of TamperSec

  2. Get funded to go full‑time.

    $100K+ to work on your organization full‑time before you raise, and introductions to the investors and philanthropic funders in this field, including our Funding Network.

    “We got our first funding because of Alexandra’s support. This was key to getting Timaeus off the ground.”

    Jesse HooglandCo‑founder of Timaeus, now Resolution · $160M+ raised

  3. Support built around you.

    1:1 coaching on your specific challenges, from people who have scaled companies themselves, like Trish Kennelly, former COO of On Deck.

    “Robocurve wouldn’t have existed if it weren’t for Hugo and the team at Catalyze Impact.”

    Jay ChooiCo‑founder of Robocurve · $10M seed

Some of the field’s top founding opportunities

  1. Hold AI companies accountableIndependent firms that check what AI companies claim about their models, the way auditors and ratings agencies do for finance.For example: An audit firm for AI modelsA ratings agency for AI companies
  2. Defend against catastrophic misuseStop AI from being used to engineer pathogens or attack critical infrastructure.For example: Early warning for new pathogensTesting the safeguards on DNA orders
  3. Keep power in checkMake sure AI doesn’t let a small group take control over everyone else.For example: A watchdog for AI in governmentAudits of who can steer an AI
  4. Make AI agreements enforceableTechnology that lets countries and companies check that the other side is keeping its AI commitments.For example: Verification built into AI chipsA verification toolkit for treaties
  5. Catch AI incidents earlyIndependent teams that detect AI-driven attacks in the real world and investigate reports of misuse, so worrying behavior doesn’t go unnoticed.For example: Detection of AI-driven attacksInvestigations of AI incidents

Ideas from Coefficient Giving’s Tailwind program (opens in a new tab), among others we’re excited about.

Know what you want to build, or still figuring it out?

Start building

Questions first? Send us a message (opens in a new tab)