It’s the sort of apocalyptic vision that has haunted scientists for years: a lone actor using artificial intelligence to engineer a pathogen combining the incubation period of HIV, the transmissibility of measles, and the fatality rate of smallpox. In a recent report, the Center for AI Safety warned that this scenario—once the stuff of science fiction—has become a credible risk in the age of generative AI.
For Kathleen McMahon, the cofounder and chief executive officer of Valthos, such risks are not abstract. Her nine-person New York–based startup is building AI software designed to detect and defend against biological threats before they spiral into catastrophe. Emerging from stealth this week with $30 million(₹263 crores) in funding from OpenAI, Founders Fund, and Lux Capital, Valthos positions itself at the intersection of biotechnology, artificial intelligence, and national security—fields now converging faster than regulators can keep up.
Inside the AI–Biosecurity Experiment
At its core, Valthos’s platform aggregates biological data from both government and commercial sources—ranging from wastewater and air-quality samples to genomic sequencing data—and uses AI models to detect anomalies and identify potential emerging threats.
The company is also building systems capable of redesigning medical countermeasures in real time, updating vaccines and therapeutics to respond to rapidly evolving biological agents.
“The only way to deter an attack is to know when it’s happening and deploy countermeasures fast,” McMahon said. She draws on years of experience leading Palantir Technologies’ life sciences division, where she learned the value of “meeting operators where they are”—a principle she now intends to apply to government biosecurity infrastructure.
Her cofounder, Tess van Stekelenburg, a Lux Capital venture partner with backgrounds in neuroscience and biology, had long sought a biodefense company to back. “This should be treated on the same threat level as nuclear or cyber,” said Lux Capital partner Brandon Reeves, who believes Valthos marks the start of a new investment wave in the field
The First Firm to Assess Your DFIR Capability Maturity and Provide DFIR as a Service (DFIRaaS)
Silicon Valley Rediscovers Defense—and the Biology of Risk
For years, venture capital largely shied away from biodefense. PitchBook data shows that 2025 saw the lowest biotech investment levels in more than a decade, even as funding poured into autonomous drones, nuclear systems, and defense software. But the rise of AI—and its dual-use potential—appears to be changing that.
“Pre-2025 there were murmurs, but now it’s really come to the forefront,” said Delian Asparouhov, partner at Founders Fund, who previously worked with Stekelenburg when Lux backed Varda Space Industries. Advances in generative AI, he said, have expanded both the ability to create bioweapons and the tools to stop them.
“Backing Valthos wouldn’t have made sense a few years ago,” Asparouhov admitted. “Now the technology makes it necessary.”
Even within OpenAI, the move signals a growing recognition of biotech’s national-security implications. Chief Strategy Officer Jason Kwon described Valthos as the company’s “first biosecurity investment,” emphasizing that safeguarding large language models against misuse will require building a broader ecosystem.
“There needs to be a system of countervailing technologies to make the system robust,” Kwon said.
The Policy Vacuum and the Global Stakes
While private capital begins to fill the gap, government funding for biodefense has been shrinking. Federal budget cuts have reduced public spending on biosecurity projects even as the threat landscape expands. “It’s hard for companies with private dollars to make it out of the investment valley of death,” warned Daniel Regan, senior fellow at the Council on Strategic Risks. “We need a pathway for these technologies to succeed.”
The National Security Commission on Emerging Biotechnology issued a stark warning earlier this year:
“There will be a ChatGPT moment for biotechnology—and if China gets there first, no matter how fast we run, we will never catch up.”
That sense of urgency now drives both investors and policymakers toward a new technological arms race—one in which AI is both weapon and shield.
For McMahon, the mission is clear but daunting.
“Biosecurity can’t be an afterthought,” she said. “We have to build systems as if the next threat is already out there.”
