AI startup ideas in Security
Powerful AI raises the ceiling on what attackers, fraudsters and careless systems can do. The most valuable security companies of the next decade will be differentially defensive: they make the world harder to break faster than they make it easier to attack.
This hub collects ideas aimed at that frontier, including machine-readable software surfaces that keep agents accountable, whistleblower infrastructure that makes fraud detection profitable, safety-first AR companions and immigration systems that are both humane and enforceable. Each thesis states how it reduces catastrophic risk, not just how it captures revenue.
5 open-sourced theses
Murmuration Engine
AI Agent Swarm Intelligence
An AGI-native strategy and execution engine that helps ambitious startups run rapid agent experiments and compound learnings.
VIEW IDEA
AURA
Marketplace for AR embedded AI Avatars
An SDK and marketplace that lets any developer drop lifelike, spatially aware AI companions into AR apps.
VIEW IDEA
Agentable
Make software legible to people and AI
Turns websites and apps into a machine-readable UI map, guides humans in-flow, and lets AI agents complete the same tasks.
VIEW IDEA
CivicPath
Immigration Dashboard
A government and immigrant-facing dashboard that makes immigration earned, legible, and enforceable through live probabilistic outcome mapping.
VIEW IDEA
WhistleWorks
Make Identifying Corruption Profitable
A privacy-first whistleblower platform that helps insiders turn fraud evidence into high-merit legal cases, then uses the resulting case data to build better fraud-detection systems.
VIEW IDEANot sure which one is yours?
Answer six questions and get matched to the ideas that fit your skills, capital and conviction.
Unsolved security problems
High-priority gaps from the Problem Atlas that still need a founder.
Antimicrobial Resistance Defense Stack
Priority 89We are walking into a post-antibiotic world, but we still operate hospitals and drug markets as if antibiotics are an infinite, cheap, reliable substrate.
Pandemic 100-Day Countermeasure Platform
Priority 88We proved we can design vaccines fast, but we cannot reliably go from novel pathogen to globally available countermeasures in a predictable time box.
AI Evaluation and Incident Infrastructure
Priority 80We are deploying high-impact AI systems faster than we can measure their failure modes in the real world.
Media Provenance Verification
Priority 78We can generate convincing audio, images, and video cheaply, but we cannot cheaply verify what is real.
Space Weather Early Warning
Priority 74We run grids, satellites, and communications on assumptions of solar calm, but severe geomagnetic storms still arrive and can cascade into infrastructure disruption.
Software Supply Chain Attestation
Priority 73We run critical infrastructure on software we cannot reliably inventory or verify, and we still treat build provenance as optional metadata.