The AI Prescription: Accelerating the End of "Incurable" Diseases

Discover how AI is accelerating drug discovery for Parkinson's, superbugs, and rare diseases, reducing screening times from months to hours in 2026.

By: AXL Media

Published: Mar 30, 2026, 9:18 AM EDT

Source: BBC new

Winning the Arms Race Against Superbugs

Humanity is currently losing the battle against antibiotic resistance, with approximately 1.1 million annual deaths attributed to untreatable infections—a figure projected to reach 8 million by 2050. AI is now reversing this trend by allowing scientists to design entirely new molecular structures that target bacteria in ways existing drugs cannot. At MIT, researchers used AI to screen 45 million compounds, successfully identifying two highly effective candidates against gonorrhoea and MRSA. These AI-designed molecules appear to bypass current bacterial defense mechanisms, offering the first real hope for a new class of antibiotics in decades.

Halting Parkinson’s Before Symptoms Begin

Parkinson’s Disease, which affects over 10 million people globally, has lacked a disease-modifying treatment for over 200 years. Current research at the University of Cambridge is using machine learning to target "Lewy bodies"—clumps of misfolded proteins that drive neurodegeneration. While traditional screening might analyze one million molecules over six months at a cost of millions, AI can now screen billions of molecules in a few days for a few thousand pounds. This efficiency has allowed scientists to identify five promising compounds that aim to stabilize proteins in their healthy state, potentially preventing the onset of the disease entirely rather than just managing tremors and stiffness.

The "Every Cure" Movement: Repurposing Existing Drugs

Treating a disease doesn't always require inventing a new molecule; often, the cure already exists in a pharmacy for a different condition. AI models at Harvard Medical School and the nonprofit Every Cure have identified nearly 8,000 approved drugs that could potentially be repurposed to treat 17,000 different diseases. This is particularly vital for the 7,000+ rare diseases that are often neglected by big pharmaceutical companies due to low profit margins. By using AI to match drug properties with disease biomarkers, researchers have found potential treatments for Idiopathic Pulmonary Fibrosis (IPF) and Wilms tumour, utilizing safe, low-cost medications that have already cleared years of safety testing.