Diagnostics

Date: 2026-07-15 (Updated) Version: 3.0 Previous Edition: 2026-06-15 (v2.0) Coverage: June 15 – July 15, 2026 catchup Catchup Update: Developments & Analysis (June 15 – July 15, 2026) Summary: The 30-day period since v2.0 saw the AI-in-medicine landscape shift from model-release frenzy to deployment, regulation, and infrastructure realities. Key themes: (1) the Fable 5/Mythos 5 saga resolved with restored access on July 1 after US export controls set a precedent for government intervention in medical-capable models; (2) OpenAI and Anthropic both launched formal partner programs and services arms — directly enabling scaled medical AI deployment through enterprise channels; (3) AI infrastructure spending reached $1.37T for 2026, but memory shortages (+74% DRAM prices) and data center protests create headwinds; (4) AI security threats escalated 89% YoY, raising stakes for HIPAA-compliant medical AI deployment; and (5) Microsoft Agent 365 reached GA with tens of millions of agents — a platform for healthcare agentic workflows. ...

Date: 2026-06-15 (Updated) Version: 2.0 Previous Edition: 2026-05-29 (v1.0) Coverage: May 29 – June 15, 2026 catchup Catchup Update: Developments & Analysis (May 29 – June 15, 2026) Summary: The 17-day period since v1.0 saw unprecedented developments affecting medical AI: Anthropic’s most capable models were launched and then government-shutdown, Chinese open-source models leapfrogged with million-token contexts, landmark AI liability rulings reshaped risk for health platforms, a major consulting scandal undermined trust in healthcare AI advice, and pharma sector AI-driven displacement accelerated dramatically. ...

Date: 2026-05-29 Status: Initial Edition Next Update: TBD Executive Summary Artificial Intelligence is transforming medicine at an accelerating pace. What was once confined to research labs and pilot studies is now becoming clinically deployed, FDA-cleared, and patient-facing. From AlphaFold 3’s atom-level biology simulations to AI copilots integrated into hospital EHR systems, the landscape in mid-2026 is markedly more mature than even 12 months prior. Key themes emerging: From diagnosis to prediction — AI is shifting from reactive diagnostics to proactive, predictive healthcare Multimodal AI — Models that combine text, imaging, genomics, and wearable data are outperforming single-modality systems Democratisation — Open-source models (DeepSeek, Qwen, Llama) are being fine-tuned for medical use, reducing costs Regulatory maturation — FDA clearance pathways for AI/ML-based SaMD (Software as a Medical Device) are now established Patient-facing AI — Major platforms (ChatGPT Health, Microsoft Copilot Health, Perplexity Health) now directly serve patients Domain 1: Drug Discovery & Development 1.1 AlphaFold 3 & Protein Biology DeepMind’s AlphaFold 3 (released 2024) represented a leap beyond previous iterations by modelling protein interactions with DNA, RNA, and small molecules — not just protein folding alone. By mid-2026, AlphaFold 3 has been integrated into pharmaceutical R&D pipelines at nearly every major drug company. ...