Biological AI Is Coming—And We’re Not Ready for Its Ethical and Global Impacts

When most of us think about artificial intelligence, we picture glowing server racks, driverless cars, or chatbots that can spin out essays in seconds. But researchers in Australia have taken the concept of AI to a far stranger place: inside living cells.

At the University of Sydney and the Centenary Institute, scientists have built a platform called PROTEUS (PROTein Evolution Using Selection). It doesn’t run on silicon chips. It runs on life itself. Inside mammalian cells, PROTEUS mutates proteins, tests their performance, and selects the best survivors—compressing a process that once took years into mere weeks. The result: molecules that are not just designed for humans in theory but evolved in human-like biology from the start.

This is not the first time biology has been reprogrammed for innovation. Frances Arnold’s Nobel Prize–winning work on “directed evolution” showed that proteins could be coaxed into useful forms through repeated cycles of mutation and selection. CRISPR gene editing, mRNA vaccines, and lab-grown organs have each pushed the boundaries further. PROTEUS, however, represents a leap: directed evolution not in bacteria or test tubes, but in the very same kind of cells therapies are meant to heal.

The implications are staggering. Imagine cancer-detecting nanobodies fine-tuned for the human immune system, CRISPR editors customized in weeks instead of years, or rapid-response vaccines that adapt alongside emerging pathogens. Unlike conventional drug discovery pipelines—slow, expensive, and prone to failure—biological AI platforms like PROTEUS promise to speed up and personalize medicine on a scale we’ve never seen.

But this isn’t just about science; it’s about power. The race for biological AI is quietly becoming geopolitical. Australia’s breakthrough puts it on the map, but China, the U.S., and Europe are all racing to dominate protein design and synthetic biology. Just as digital AI has become a battlefield of patents, data, and regulation, biological AI could become the next strategic technology defining national advantage. Whoever leads in this domain won’t just shape drug markets; they’ll shape the very architecture of human health.

And then comes the economics. Biotech is already dominated by a few global giants, with eye-watering costs for new drugs that often put them out of reach for patients in poorer countries. Will biological AI democratize drug discovery—allowing smaller labs and even start-ups to evolve bespoke therapies—or will it cement the power of those who already control the pipelines? In a world where access to medicine is one of the starkest inequalities, the distribution of this technology matters as much as its scientific promise.

Of course, governance will be crucial. Unlike digital AI, which can be switched off with a keystroke, biological systems are messy, self-replicating, and unpredictable. A protein optimized inside mammalian cells might behave beautifully in the lab but trigger unforeseen consequences in the body. How do you regulate something that keeps evolving? Should the guardrails for biological AI be stricter than those we apply to digital AI? At minimum, global oversight mechanisms will need to emerge before biological AI scales into medicine.

Then there is public trust. Science communication has always been fraught, from GMOs to lab-grown meat to vaccines. Telling the public that scientists have built an “AI inside your cells” risks inflaming fear unless the metaphor is explained carefully. PROTEUS doesn’t think or feel. It isn’t a living chatbot. The AI analogy is metaphorical: it describes a process of self-optimization, where millions of variations compete and the best survive. Still, metaphors matter. If scientists aren’t careful, “biological AI” could trigger the same backlash that GMOs once did—more politics than science.

And finally, the philosophy. The very phrase “biological AI” forces us to question what counts as “artificial.” If a system learns and adapts using DNA and proteins rather than code and silicon, is it still “artificial”? Or have we blurred the line between human design and natural evolution? For centuries, intelligence has been the hallmark of human uniqueness. First machines imitated it in algorithms; now biology itself is being recruited to solve problems. Intelligence, it seems, is no longer the sole province of minds—or even of machines.

The age of digital AI has already reshaped our societies, often faster than we could regulate. The age of biological AI could reshape life itself. And this time, the stakes are higher: not just data, but our own biology.

Australia’s PROTEUS system is a milestone. But whether it ushers in a future of democratized medicine or deepened inequality, of health breakthroughs or ethical crises, will depend less on what scientists can do and more on what societies decide they should.

We once feared machines might outthink us. Now we must ask: what happens when biology itself learns to out-evolve us?