Artificial intelligence is rapidly transforming the biological sciences, creating both significant opportunities and profound risks for global health security. While AI can empower malicious actors to design biological weapons with unprecedented ease, it also offers powerful tools for tracking disease outbreaks and delivering critical public health information, potentially stopping deadly contagions in their tracks.
This dual nature of AI has ignited a critical race between offensive and defensive capabilities in the biological sphere. Experts are increasingly concerned about the proliferation of 'dark biology', where dangerous pathogens could be engineered in secret, contrasting with the promise of a new era of collective global health bolstered by AI's analytical power.
Every infectious disease outbreak begins small, and public health success hinges on early detection before widespread transmission overwhelms healthcare systems. AI's greatest asset in this context is its ability to drastically reduce the time between an outbreak's emergence and its detection. This speed can be the deciding factor between containing a localised outbreak and facing a full-blown epidemic or even another global pandemic.
The World Health Organization (WHO) is already harnessing AI for this purpose. At its Hub for Pandemic and Epidemic Intelligence in Berlin, computational systems utilise AI and pathogenic surveillance to monitor infectious disease outbreaks worldwide. Central to this effort is a system called Epidemic Intelligence from Open Sources (Eios), which acts as a vast filter for real-time global health information. Unlike traditional methods that rely on official case counts, which can take days or weeks to compile, Eios leverages modern computing and social media to gain an immediate picture of emerging threats. It scans millions of websites, social media platforms, and local news reports in multiple languages, filtering out 'noise' to identify genuine signals of an outbreak.
The offensive potential of AI in biology is no longer theoretical. AI systems can guide researchers through immense bodies of biological knowledge, explaining complex laboratory techniques, locating obscure scientific literature, and even suggesting experimental designs. While these capabilities are not inherently dangerous, they significantly lower the barrier to entry for what was once a highly specialised field. The most significant concern is AI's potential to assist in creating genuinely novel biological forms – organisms with capabilities not found in nature. Unlike a software bug that crashes a computer, a biological mistake or weapon, once released, cannot simply be switched off or recalled, as living organisms can reproduce, spread, mutate, and evolve independently.