LONDON – AI could bring humanity closer to finding treatments for several life-threatening illnesses like cancer, and while scientists question the timeline of AI giants, clinical studies and lab breakthroughs suggest AI is already making measurable progress in the fight against cancer.
Anthropic CEO Dario Amodei made striking prediction about the future of AI in medicine. Amodei believes machine intelligence could revolutionise medical science so dramatically that some of humanity’s most devastating illnesses may become largely obsolete within the next ten years.
Sharing his thoughts at Anthropic’s 2026 initial public offering (IPO) roadshow, the executive claimed that cancer, Alzheimer’s, heart disease and mental illness could become “things of the past” in less than a decade. His prediction sparked intense debate across the technology and healthcare industries, with some researchers share remarkable progress in AI-assisted cancer detection and experimental treatment development, while others warn that curing complex diseases is far more difficult than accelerating computer-based research.
Amodei’s extraordinary confidence stems from his belief that advanced models could accelerate biological research and drug discovery by as much as tenfold. He said scientific advances which would normally take 50 to 100 years could potentially be achieved within just 5-10 years.
Such progress, if realised, could transform how scientists see diseases, discover drugs and develop personalised treatments. Other prominent technology executives, including Google DeepMind CEO Demis Hassabis and Elon Musk, have also expressed optimism about AI’s potential to transform healthcare.
One major breakthrough supporting this optimism is Google DeepMind’s AlphaFold system, which has predicted the three-dimensional structures of more than 200 million proteins. These predictions have given researchers powerful new tools for studying disease mechanisms and identifying potential drug targets.
Although Amodei’s ten-year prediction remains unproven, scientific research produced increasingly compelling evidence that AI can improve certain aspects of cancer diagnosis and experimental treatment development.
AI vs Breast Cancer
One of the clearest demonstrations of AI’s medical potential comes from Sweden, where a large clinical trial involving approximately 106,000 women examined whether artificial intelligence could improve breast cancer screening.
Mammography Screening with Artificial Intelligence (MASAI) trial found that AI-supported screening detected 338 cancers among 53,043 women, compared with 262 cancers among 52,872 women undergoing conventional screening. The findings shows roughly 29% relative improvement in the cancer detection rate, while reducing radiologists’ image-reading workload by approximately 44%.
AI vs Colon Growths
AI has also demonstrated considerable potential in detecting abnormalities associated with colorectal cancer. A large analysis of 46 randomised controlled trials involving more than 37,000 participants found that AI-assisted colonoscopy improved the detection rate of adenomas, abnormal growths that can sometimes develop into cancer.
The analysis reported approximately 22% higher adenoma detection compared with conventional colonoscopy. AI-supported systems can examine live colonoscopy footage, identify suspicious abnormalities and alert doctors to growths that might otherwise be overlooked.
AI vs Immune Cells Therapy
Perhaps the most relevant development to Amodei’s prediction comes from Anthropic’s own biotechnology collaborations. Working with Generate Biomedicines, researchers used Anthropic’s Claude AI to redesign a component of CAR T-cell therapy, a treatment that modifies patients’ immune cells to recognise and attack cancer.
AI-assisted design reduced the size of a key genetic component by approximately 70%. During laboratory testing, the redesigned system enabled immune cells to produce roughly ten times as many cancer-recognising receptors and showed improved tumour-cell killing activity.
CAR T-cell therapies have already demonstrated success against certain blood cancers. Researchers are now investigating whether improved designs could help extend these treatments to solid tumours. However, the new AI-assisted designs remain experimental and have not yet demonstrated safety or effectiveness in human patients.
AI vs Antibodies
Another Anthropic collaboration, involving BigHat Biosciences, has produced promising laboratory results in antibody engineering. Antibodies can be designed to recognise proteins associated with cancer cells, potentially helping deliver medicines directly to tumours or guiding immune responses.
Scientists working with Claude tested approximately 1,010 molecular variants over ten weeks. The collaboration produced improved antibody candidates, with some exhibiting more than tenfold stronger target binding and improved stability under laboratory conditions.
