Journal

Follow one AI tool that prioritises head CT scans through EU, UK and US device rules, the AI Act timetable and routes to compensation after harm.

A shared clinical AI interface can contain functions with different purposes and risks. Tandem's expanding suite shows why product boundaries and change control matter.

Why model updates, retrieval changes and personalisation challenge one-off assessments, and what proportionate lifecycle oversight could look like.

How to compare intended purpose, classification, evidence, market access and lifecycle costs after the UK decision on indefinite CE recognition.

Doximity's public terms and clinical product messaging raise a precise intended-use question. Read the documents without jumping to legal conclusions.

The AI harness is the software around a model. Learn how instructions, context, retrieval, checks and permissions shape a clinical product, and what buyers should evaluate.

Evaluate healthcare growth platforms using appropriate completed care, incremental contribution, clinician time and patient experience rather than traffic or booking volume alone.

Calculate the combined cost of practice subscriptions, discovery fees, payment processing and administration, using a transparent illustrative clinic model.

Explore how patient-discovery platforms are expanding into AI reception and practice workflows, and what that could mean for clinicians' costs, control and patient access.

Assess what US mental-health clinicians gain and give up when using platforms for insurance access, billing and practice operations instead of arranging each function independently...

Decide whether an Australian GP practice needs patient discovery, a digital front desk or both, using capacity, continuity, workflow and cost as the starting points.

A clinician-led evaluation of Zocdoc's Zo AI phone assistant, including scheduling accuracy, escalation, privacy, human handover and a controlled pilot.