If you tell ChatGPT you prefer bullet points and Claude you’re planning a trip to Tokyo, neither assistant remembers the other’s conversation. That’s the fragmented state of personal AI context today, and it’s the precise wedge Egoist Machines is trying to pry open. The San Francisco startup, founded in 2026, is building what it calls an AI Passport, a user-controlled layer meant to make personal data portable across major AI assistants [Y Combinator, October 2026]. The bet is that as people use more AI tools, they’ll want a single, secure profile for their preferences, history, and goals,not a dozen separate silos.
It’s an infrastructure play with a consumer face. The product, launched alongside a $3 million seed round in October 2026, works like a permissions manager. A user maintains one portable profile and authorizes applications to access selected slices of information, for a defined period, with a receipt generated when data is used [Cambridge St Catharine’s College, August 2026]. The company likens it to a Google login, but with granular control over what is shared [Henry Li - A* AI | LinkedIn, 2026]. The go-to-market strategy is twofold: distribute directly to consumers through first-party products, and integrate the Passport into existing consumer applications that already have distribution [Y Combinator, Unknown].
The technical and social wedge
The product’s differentiation rests on two pillars: technical interoperability and a stated commitment to user sovereignty. On paper, AI Passport is designed to work across assistants like ChatGPT, Claude, Gemini, and Muse, acting as a neutral conduit [Pulse 2, October 2026]. The more ambitious claim is the social contract. By framing the product as a “user-owned memory layer,” Egoist Machines is appealing to growing unease about data lock-in and opaque usage within walled AI gardens [The Founder's Pack, 2026]. The promise of receipts and time-bound access is a direct response to that anxiety.
For now, traction is measured in pilots, not production contracts. The company announced in October 2026 that it had more than 10 business-to-business-to-consumer pilot programs underway, each involving organizations with over 1 million users [Pulse 2, October 2026]. Those organizations were not named, and the nature of the pilots,whether they are technical integrations or purely exploratory,remains unspecified. This is a common early-stage signal, but the path from pilot to paid integration, especially at the consumer application layer, is long and fraught.
A founding team built for narrative
The company’s story is powered by a founder duo with complementary, headline-friendly backgrounds. Co-founder and CEO Erin Meryl McGurk is described as having grown a social media following from zero to 1 million in one year and is the founder of the Girl Economics newsletter [Y Combinator, October 2026]. Co-founder and CTO Dr. David Khachaturov brings the technical credibility: a PhD in Computer Science from Cambridge, former ML research work at OpenAI on the PaperBench project, and AI security research experience [Y Combinator, October 2026]. The pairing seems engineered to speak to both consumer adoption and deep technical feasibility.
| Founder | Role | Key Background |
|---|---|---|
| Erin Meryl McGurk | Co-founder & CEO | Grew social media to 1M followers in one year; founder of Girl Economics newsletter [Y Combinator, October 2026]. |
| Dr. David Khachaturov | Co-founder & CTO | PhD in Computer Science, Cambridge; former ML Researcher at OpenAI; AI security background [Y Combinator, October 2026]. |
The seed round and its signaling
The $3 million seed round closed in October 2026 provides the initial fuel. The investor list is notable for its blend of institutional and technical validation: Y Combinator, Goodwater Capital, a European family office, and individual scientists from DeepMind and Meta participated [Yahoo Finance, October 2026]. The involvement of AI researchers from leading labs is a strong signal of technical interest, if not a commercial endorsement. The capital will presumably fund the team’s expansion from its current two-person size and the development of both the core infrastructure and initial consumer-facing products [Y Combinator, Unknown].
The adoption funnel ahead
The most significant risk for Egoist Machines isn’t technical feasibility; it’s the classic two-sided marketplace problem. The value for users is only realized when their AI Passport works across multiple assistants and apps they care about. The value for application developers, however, only materializes when a critical mass of users have and actively use a Passport. Breaking this cycle requires either spectacularly successful first-party products that create a user base, or signing partnership deals with major distribution platforms willing to integrate early.
The company’s early focus appears to be on the latter, through its B2B2C pilot programs. The realistic competitive set, however, isn’t other startups. It’s the platforms themselves deciding to build similar functionality in-house, or simply deciding that user lock-in is more valuable than interoperability. Apple’s Keychain or Google’s password manager are analogous in concept but owned by the ecosystem giants. For AI Passport to succeed, it must convince both users and developers that a neutral, user-owned layer is preferable to one controlled by any single AI provider.
The ideal customer profile here is not an individual consumer, but the product lead at a popular consumer-facing app,a note-taking tool, a travel planner, or a fitness coach,who is looking to integrate AI features without forcing users to start from scratch every time. For them, Egoist Machines is selling a faster path to personalized AI and a privacy narrative they can market. The next twelve months will be about converting those unnamed million-user pilot organizations into launch partners with public integrations. If they can land one, the narrative shifts from a interesting protocol to a viable bridge between the AI islands.
Sources
- [Y Combinator, October 2026] Launch YC: Egoist Machines: user-owned hyper-personalisation infrastructure for Web 4.0 | https://www.ycombinator.com/launches/St3-egoist-machines-user-owned-hyper-personalisation-infrastructure-for-web-4-0
- [Cambridge St Catharine’s College, August 2026] Catz startups turning research to real-world benefit | https://www.caths.cam.ac.uk/startups
- [Pulse 2, October 2026] Egoist Machines Raises $3 Million Seed Round And Launches AI Passport | https://pulse2.com/egoist-machines-raises-3-million-seed-round-and-launches-ai-passport/amp/
- [Yahoo Finance, October 2026] Egoist Machines AI Passport Goes Live Today | https://finance.yahoo.com/technology/ai/articles/egoist-machines-ai-passport-goes-132300417.html
- [Henry Li - A* AI | LinkedIn, 2026] Post on AI Passport functionality | Source snippet
- [The Founder's Pack, 2026] Article referencing Egoist Machines | Source snippet
- [Y Combinator, Unknown] Egoist Machines company profile | https://www.ycombinator.com/companies/egoist-machines