Liaise helps close the generalist talent gap in AI safety by helping motivated non‑technical people enter with fluency, alignment, and a trusted network.
Liaise helps close the generalist talent gap in AI safety by helping motivated non‑technical people enter with fluency, alignment, and a trusted network.
Project Details
Updated 08/05/26 · Edited by orgWe are currently running Liaise (https://liaiseai.xyz/) as a volunteer-driven project. It is a structured bridge for generalists entering AI safety.
London has a huge pool of talented operations people working in elite finance and consulting firms along with other generalist fields. These people are less exposed to AI safety as it stands and as a result are less certain on how they could have an impact on the problems. However, they are exactly the sorts of people who understand how to improve operations and work well as generalists, as their professional experience gears them up to making projects run well. This matters because of the well discussed gap in generalists as opposed to researchers.
AI safety is not just a research problem anymore. It’s also a coordination, governance, and implementation problem. The field needs generalists, communicators, organisers, builders, and connectors to help turn research into action and bring more people into the ecosystem. We need this generalist talent to make sure AI goes well.
We recently organised the generalists mixer in central London - https://luma.com/3bc7capv
Over the next 6 months - We will be working on
-
Regular mixers and easy networking for generalists exploring their fit in London. The baseline is people with at least some previous work skills and who care about AI risk but don’t yet have fluency or a network.
2 events delivered so far: -
Pub mixer recap: https://www.linkedin.com/posts/liaiseorg_mixer-aisafety-liaise-activity-7483898130560647169-6cx9
-
Picnic mixer recap: https://www.linkedin.com/feed/update/urn:li:activity:7490740308520030208
-
Two generalist‑focused hackathons: One of the hackathons is currently WIP; a one-day hackathon with Frame fellowship in London (end of July 2026) to bring content creators and communicators in the space
-
Design and get feedback on the LI(ai)SE incubator fellowship: a short, cohort‑based programme where a small group of generalists get structured content, a scoped project, and support applying to roles or programmes or starting their own projects to help AI go well.
We currently have a team of 5 volunteers helping organise the work in London and have 5 advisors from EA and AI Safety adjacent organisations. (https://liaiseai.xyz/#page-2)
Right now I’m the lead for running the project (design, website, outreach, logistics, facilitation, comms), with a small set of advisors and core team volunteers along with AIS organisations offering mentoring, feedback, and occasional co‑hosting.
We are raising our first funding to cover ops (team+working) costs along with the costs of organising the outputs. The concrete outputs are:
- 2-3/per month low friction mixers and events for people to test their context and ecosystem understanding, (thematic and general mixers)
- at least one hackathon a month with real outputs produced, \
- a 3-month documented fellowship pilot (including curriculum and example projects),
- and a small but growing group of generalists who can point to LI(ai)SE as part of how they became legible candidates for non‑technical AI safety roles.
Team's track:
Rohan, Co founder
- Ex policy and programs for Science & Tech office in Government of India
- BlueDot Impact AGI Strategy & AI Governance courses
- Arcadia Gov Fellowship
- Raised rapid grant through BlueDot to run Liaise ops
- Built Liaise's online presence of 2k+ on LinkedIn in 2 months
Look at our team profile here: https://liaiseai.xyz/
Our advisors - Elizabeth Pasatembou Researcher in Quantum Communication & Networks Founding Team/Advisor (Liaise)
- Gergő Gáspár Director Effective Altruism UK
- Moneer Moukaddem Career Advisor Successif
- Joshua Landes Head of Community & Events, BlueDot Impact
Theory of Impact
Updated 08/05/26 · By grantmaking.aiLi(ai)se reduces x‑risk from AI by fixing a specific bottleneck: the shortage of high context generalists in non‑technical AI safety roles (ops, comms, governance, policy‑ops, fieldbuilding). If we can reliably turn 'motivated, slightly lost generalist' into credible, well‑networked candidate with real work under their belt, AI safety orgs can hire and scale faster, and existing technical and policy work has a much better chance of being implemented well.
In practice, LI(ai)SE does three things: it gives people basic fluency in AI safety and governance, helps them ship a first concrete project (e.g. ops improvement, comms product, governance memo), and connects them with mentors and hiring managers who can vouch for them. That combination makes it more likely that they end up running evaluations, incident processes, fieldbuilding programmes, and governance projects at the orgs that matter.
Over time, having more of these people in the right places strengthens the institutional and operational “scaffolding” around advanced AI systems, which is critical for actually preventing and responding to catastrophic failures, not just talking about them. If LI(ai)SE succeeds, more aligned generalists end up in key non‑technical roles at AI safety orgs, which makes those orgs better at actually implementing safety, governance, and fieldbuilding work that reduces x‑risk from advanced AI.
People
Updated 08/05/26 · By grantmaking.aiTeam Member
Funding Details
- -
- -
- 1 year
- -
- -
- -
- -
- -
- funded for events partially by rapid grants
- -
Track Record
What have we built:
- Liaise slack community of 150+ active users looking to move into AI Safety space as generalists + 10 mentors from AIS fieldbuilding orgs An in-person generalists mixer in central London with 30+ people from diverse backgrounds pivoting into AIS - https://luma.com/3bc7capv
- LinkedIn presence: 2k+
- 3 core team members pivoting to AI Safety within the first month of joining the volunteer core team
- CRM of high-impact professionals with 8+ years of experience and matching them with mentors and orgs
- A WIP document of 'template for how to pivot into AI safety as a generalist'
- Designing a hackathon with Frame Fellowship on content and communications for AI Safety
- Building a database of who gets hired most often across top AI safety orgs, and what those people’s backgrounds look like. It’s a way to make the field easier to navigate for candidates and hiring teams
- A core incubator fellowship for people to test their ideas/researc/projects with mentor support and funding support - to find their fit in AIS. (expediting alignment, fluency and placements).
Discussion
I think Rohan is great, we need more AIS projects happening in London!
Thanks @Gergő Gáspár for your mentorship and advice!
I'm a generalist who's now a part of LI(ai)SE's growing AI safety community!
thanks for your comment and support to the community @Ammara Durrani