Artificial intelligence is beginning to reshape every stage of scientific research and healthcare delivery. For Katherine Chou, VP and Head of Product for Google Research, its greatest promise in precision medicine lies in discovering new biomarkers, developing targeted therapies and tailoring care to the needs of each individual.
Recent advances in AI are also changing how researchers generate hypotheses, analyse complex evidence and collaborate with specialist experts. More interactive and self-improving systems could accelerate scientific discovery while helping research teams explore relationships between diseases, biological signals and treatment responses that were previously difficult to identify.
A major opportunity comes from multimodal AI, which can connect previously fragmented sources of information, including genomics, clinical records, medical imaging, pathology, wearable devices and real-world evidence. By analysing these signals together, AI systems may identify health patterns, emerging risks and biological subtypes that would be difficult to detect through any single dataset.
This could lead to more accurate risk assessments, earlier interventions and treatments designed around a patient’s specific biology, environment and lifestyle. Continuous information from wearable devices could also be combined with relatively stable genetic data to reveal changes in a person’s health trajectory over time. Rather than relying on broad, uniform treatment pathways, clinicians could make decisions informed by a more complete and dynamic picture of each patient.
Multimodal systems may also help connect precision medicine with holistic care. Factors such as behaviour, living conditions and environmental exposure can have a substantial influence on health outcomes. AI could act as a translation layer between these social and environmental factors, a patient’s genetics and the clinical outcomes that healthcare teams are trying to achieve.
However, moving AI from research into clinical practice requires more than an accurate model. Systems must protect privacy, operate securely and fit naturally into clinical and patient workflows. They should reduce administrative burden, improve access and demonstrate measurable benefits to health outcomes. At a wider level, they must also align with regulation, public policy and responsible health economics.
Trust will be essential. Generative AI can be powerful, but its outputs must remain grounded in evidence, independently validated, verifiable and open to human review. Scientists and clinicians should use AI to strengthen professional judgement, not replace it or encourage over-reliance on automated recommendations.
Success would mean broadly adopted AI tools that accelerate scientific discovery while helping patients feel genuinely understood and better supported. By combining reliable medical evidence with highly personalised guidance, AI-powered precision medicine could make healthcare more preventive, responsive, accessible and human-centred.
