In Lewis Carroll's Through the Looking-Glass, the Red Queen tells Alice that it takes all the running you can do just to stay in the same place. Biologists borrowed the image decades ago to describe the evolutionary arms race between hosts and pathogens. A young biotechnology company has now taken the name for a more literal race: using artificial intelligence to build defences against viruses before they appear.
Red Queen Bio, an AI biosecurity start-up, is designing antibodies intended to protect against a range of potential pathogens, including, as The Wall Street Journal reported on 28 September 2026, the once-theoretical threat of a virus designed with the help of AI.
Who is behind it
Red Queen Bio was spun out of HelixNano, a biotechnology company, in 2025. It is led by chief executive Nikolai Eroshenko, with Hannu Rajaniemi, a HelixNano co-founder who is also known internationally as a science-fiction author, serving as executive director.
The company has raised $36 million in total. Its $15 million seed round, announced in November 2025, was led by OpenAI. Other investors include Cerberus Ventures, Fifty Years and Halcyon Futures.
OpenAI's role as lead investor is notable. The company behind ChatGPT has repeatedly warned that increasingly capable AI models could lower the barriers to creating biological weapons, and it has introduced safeguards to prevent its systems from assisting with dangerous biology. Backing a company that uses AI to strengthen defences is the other side of that effort.
How the approach works
Red Queen Bio's method is an iterative loop between computation and the laboratory. It uses AI models to design antibody candidates on computers. It then grows cells in the lab to produce and test those candidates, and feeds the experimental results back into its models to improve the next round of designs.
That cycle, often described as "design, build, test, learn", has become central to AI-driven drug discovery. Each iteration improves the models' understanding of which molecular features produce the desired effect, allowing the company to move faster than traditional antibody discovery, which relies heavily on screening large numbers of candidates.
Eroshenko has emphasised the importance of binding strength, the tightness with which an antibody attaches to its target on a virus.
"The tighter they bind, the less of them you need. Suddenly, you have something that can scale," he said.
The point is practical. Antibody therapies are expensive to manufacture. If a more potent antibody can achieve the same protective effect at a much smaller dose, the number of people who can be treated from the same production capacity rises dramatically. In a pandemic, that difference could determine whether a treatment reaches millions of people or only a few thousand.
From bird flu to smallpox relatives
The company is starting with bird flu and other influenza viruses. Highly pathogenic avian influenza has spread widely among birds and some mammal species in recent years, and public health authorities have warned of the risk that it could adapt to spread efficiently between humans.
Red Queen Bio plans to expand to coronaviruses, Ebola-like viruses and relatives of the smallpox virus. Those families have caused, or have the potential to cause, some of the most severe outbreaks in human history. The company is aiming to begin its first clinical trials in 2027.
Safety commitments
