On "human in the loop" and automation bias
A widely circulated exchange about AI and military targeting raises a question every QARA professional should sit with: does "a human makes the final decision" really resolve anything when review time is a glance? Plus what's missing from the EU AI Act's high-risk categories, and new Commission guidance for health AI.
FDA's first AI-misuse warning letter wasn’t so much about AI
Purolea Cosmetics Lab's FDA warning letter made headlines as the first to cite AI misuse in manufacturing. But strip away the AI angle and the real story is a quality unit that had already collapsed. Part 3 closes a series on why AI regulation isn't leveling the playing field — yet.
The EU AI Act's postponement rewards those who never started
The EU AI Act's high-risk deadline just moved to December 2027. Under Article 111's grandfather clause, companies that never started compliance may get a permanent pass, while those who invested early carry the cost. Part 2 of a series on the disparity AI regulation is creating.
LLMs for health advice get separate treatment
Over 5% of all ChatGPT messages are health-related, and 40 million people ask it health questions daily — yet it isn't regulated as a medical device, while purpose-built health apps are. Part 1 of a three-part series on why AI regulation creates disparity, not just innovation.
Quantum: The Regulatory Frontier That Will Catch Us Off Guard
Quantum computing is revolutionizing medical device development, but regulatory frameworks aren't ready. Companies like Algorithmiq are achieving 100x precision improvements in cancer therapy using quantum-generated synthetic data—yet FDA's 2025 AI/ML guidance doesn't address quantum validation challenges. How do you validate data you can't reproduce classically? Regulatory science needs quantum-aware frameworks now, before quantum-AI medical devices reach the clinic.
