Corma raises $60M seed to build a defensive cybersecurity AI model
What's the deal? CormaDealroom has a profile for this one. Try Dealroom →, a frontier AI lab focused on defensive cybersecurity, has raised $60 million in seed funding led by Sequoia Capital, with Khosla Ventures and Coatue participating. It is building what it calls the first foundation model purpose-built for defensive cybersecurity.
Why now? Foundation models from OpenAI, Anthropic, and GoogleDealroom has a profile for this one. Try Dealroom → have grown adept at coding and software reasoning — skills that map directly onto offensive security. Corma argues those same models fall short on defense, which demands sifting through security data, correlating weak signals over long time horizons, and staying consistent across thousands of decisions.
The evidence: Corma ran hundreds of simulations in realistic enterprise environments modeled on Fortune 500 organizations. Using models such as OpenAI's GPT and Anthropic's Claude, it found AI attackers succeeded 88% of the time, while AI defenders detected just 12%.
What's the endgame? Corma's foundation model powers AI agents that organizations onboard much like a new team member. Once deployed, the agents work across defensive security tasks, continuously learn their environment, and scale.
"The race to general intelligence in cybersecurity has already begun, and the attackers have a significant head start," said Alon Pluda, co-founder and chief executive officer of Corma. He said the goal is "to make sure the defenders win this race."
The signal: At $60 million, Corma's seed round sits in the top 1% of seed deals in its geography and industry, a sign of how aggressively investors are backing AI-native security bets. As AI sharpens attackers' speed and reach, capital is flowing toward founders promising to close the widening gap on defense.
