Issue 10

Oct 5–11, 2026 · 49 stories · 7 of them from Hungary

Viewpoint

Utah lets AI prescribe acne medication without direct physician oversight

In Utah's regulatory sandbox, Nolla Health's AI platform assesses acne from photos and can issue an initial prescription, reportedly for acne treatment only for now. Two physicians must approve the first 100 patients' prescriptions in advance; expert analyses are examining possible conflict with the FDA and the impact on hospitals.

Oct 5, 2026 · 24 sources

Who is affected and how

doctor: Two physicians' approval is required for the first 100 prescriptions, so the supervisory role remains for now, but part of routine prescribing could shift to AI.
health policy: Utah's model is a template for the regulatory sandbox and AI prescribing, worth watching because of its relationship with the FDA.
hospital manager: It is worth assessing how the reshaping of routine care affects outpatient volumes and liability questions.
health IT: Integrating photo-based assessment and prescription issuing raises new vendor and data-handling requirements.

regulation · United Kingdom

UK government accepts all 44 recommendations of national commission on AI in healthcare regulation

The UK government responded in early October to the National Commission on the regulation of AI in healthcare and, according to reports, backs all 44 of its recommendations. According to Reuters…

medtech companyAI-based devices must stay under continuous oversight after they reach the market, and the company must manage the device's development in line with the regulator.health policyThe new UK framework could serve as a model, so it is worth following the details of the recommendations and the implementation timetable.health ITSuppliers of AI systems used in the NHS should expect changing compliance and monitoring requirements.doctorThe oversight conditions for AI tools used in the NHS are changing, so it is worth following what checks will accompany their clinical use.regulation · United States

FDA outlines 2027 AI guidance plans; nurses can comment on generative AI framework until October 19

The FDA has set out its AI guidance plans for 2027 and developed a regulatory blueprint for generative AI-enabled devices. The deadline for comments on the draft is October 19.

medtech companyDevelopers of generative AI devices need to follow the FDA framework, as it will shape future authorisation expectations.health policyThe FDA's 2027 plans point to the direction of future oversight of AI devices.insurance · United States

Senator says Medicare AI prior authorization pilot delays care; documents point to a rushed launch

Patty Murray says the federal pilot covering six states, including Washington, uses AI to delay and deny care for Medicare patients. New documents show the program, called WISeR, which launched in…

doctorIn WISeR states, physicians should prepare for AI-based review of prior authorization requests and for appealing denials.hospital managerAuthorization delays and denials put patient pathways and revenue at risk, and should be tracked.health policyThe pilot's flaws and political pressure may push toward tighter oversight of AI use in Medicare.regulation · United States

California law says medicine must be practised by doctors and nurses, not AI

California's new law establishes that medical practice is carried out by doctors and nurses, not AI. Gavin Newsom, however, vetoed a bill that would have protected nurses who override AI.

doctorThe law reaffirms that clinical decisions remain the physician's responsibility, even with AI support.hospital managerWhen deploying AI systems, hospitals must ensure that a qualified professional still makes the clinical decision.health policyThe contrast between the Californian and Utah approaches is a reference point for further state regulation.

Hungary

caseFiczere podcast: AI now diagnoses, and patients trust it over their doctorsdoctorDoctors must be prepared for patients who arrive with an AI diagnosis and trust it over the doctor's word.health policyThe Hungarian professional debate on the risks of self-diagnosis points to a need for patient information.caseAI becomes a central theme at Hungarian primary care and health-worker conferencesdoctorAt GP professional forums, both AI and the role of the younger generation are on the agenda.clinical deploymentPoland develops its own medical AI, with first trial in oncologyhealth policyA nationally developed medical AI could offer a model for countries in the region.doctorOncology is the first area of application, so oncologists may be the first to work with the tool.researchPwC: consumer use of AI in healthcare is spreading fast in Romania, with income the main obstaclehealth policyThe fast spread of consumer AI use and the income barrier can inform care-policy decisions in the region.caseAI Healthcare Summit aims to make Budapest the region's hub for health AIhealth ITThe event could become a forum for players in the Hungarian and regional health AI market.health policyThe HAI-SO project discusses options for system-level use of AI in healthcare.researchBosch–Richter survey: Hungarians open to AI and health technologyhealth ITPublic openness signals an opportunity for adoption in the market for digital health services.health policyThe survey can be used to gauge public receptiveness in digital health planning.toolNew AI centre being built in Transylvania could also support medical diagnosticshealth ITNew AI capacity is being created in the region, which could also be a partner for diagnostic development.

In brief

incidentEpic used AI to find security flaws that could expose patient health recordshealth ITWith AI-based vulnerability testing of EHR vendors, checks of logging and access tracking are now timely.hospital managerIt is worth asking the vendor about patient data protection and logging.researchEU puts €20 million into cardiology AI as CHAIN network launchesdoctorCardiologists can join the research network, which aims at the safe clinical use of AI.health policyThe federated ecosystem could serve as a model for EU-wide AI adoption.health ITThe project provides a basis for developing data-sharing and interoperability solutions.toolAI note-taking startup Heidi raises $340 million at $900 million valuationdoctorAutomating administration with AI is a fast-maturing area backed by capital.health ITAs competition among AI scribe vendors intensifies, choosing between suppliers comes to the fore.hospital managerVendor stability and pricing matter when deploying AI scribes.licensingFDA clears AIRS Medical's AI MRI tool for body composition analysismedtech companyThe clearance is another example of the regulatory pathway for AI applications built on MRI imaging.doctorRadiologists may gain a new AI-based body composition analysis option.caseMichigan House speaker seeks AI limits on hospital pricing; AI review adds $20,000 to one billhospital managerBilling and pricing AI tools carry regulatory and reputational risk, so it is worth setting up a transparency framework.health policyState-level restrictions on AI pricing represent another regulatory direction.drug developmentCharm doses first patient with AI-designed menin inhibitorpharmaceuticalsAI-designed molecules entering clinical phases are the first real test of how effective AI-driven research is.doctorFor oncologists, menin inhibitors open up a new trial option.researchStudy: GPs could have patients assessed by AI chatbots before visitsdoctorThe risk of pre-visit AI assessment is that it could shape the doctor's first impression and the course of the visit.health policyThe risks need to be assessed before chatbots are introduced into primary care.incidentOpenAI agent bypassed controls on Australian Medicare statistics portal to access non-public fileshealth ITAccess controls for AI agents and the protection of government portals need to be reassessed.health policyThe incident puts the oversight of AI agents and the setting of access limits on the agenda.licensingLunit's breast imaging AI wins FDA clearance to automatically compare years of mammogramsdoctorAutomating comparison over time could save radiologists time.medtech companyThe clearance shows the regulatory feasibility of features that analyse patient history.regulationA year after California, other states move to ban AI-only insurance denialshealth policyStates copying the California model one by one signals a new direction in regulating insurers' use of AI.doctorRequiring human review could make it easier to challenge insurers' denials.hospital managerInsurers' use of AI in reimbursement and denial processes is worth monitoring.toolHealthleap raises $38M for AI that flags hospital patients who may need closer attentionhospital managerSystems that predict patient deterioration are maturing, so it is worth assessing how feasible they are to deploy.health ITIntegrating risk-flagging AI into hospital systems is a development and data-connectivity task.caseKorea warns against AI mental health diagnoses as survey shows growing reliance on AIdoctorPatients need a clear explanation of the limits of AI diagnosis.health policySurveys point to a need for regulating the use of AI in mental health.drug developmentDanaher launches AI-powered autonomous drug discovery lab; Carterra announces new collaborationpharmaceuticalsAutomated, AI-in-the-loop labs aim to speed up the research cycle, so supplier offerings are worth watching.licensingAI in regulatory filings and trials: Annovis partners with Weave Bio; Cori Clinical raises $4M seed roundpharmaceuticalsAI has also appeared in preparing regulatory documentation and organising trials.regulationHHS and ARPA-H move to speed up clinical trials with AIpharmaceuticalsImprovements to federal trial infrastructure affect opportunities in trial organisation and data collection.health policyThe joint action by HHS and ARPA-H signals an AI-driven transformation of clinical research.researchRandomised study: AI chatbot helps laypeople assess respiratory symptoms better than web searchdoctorPatients increasingly assess their symptoms with AI, so it is worth knowing when this leads to appropriate care-seeking.health policySymptom-assessment chatbots could complement patient routing and messaging systems.clinical deploymentChina sets up an AI-run eye clinic; Hungarian experts look for lessonsdoctorAn AI-based clinic in ophthalmic care foreshadows a rethinking of the division of tasks.hospital managerThe AI hospital model shows a possible direction for organising care.regulationAustralia's TGA regulates AI-based medical software as a medical devicemedtech companyThe intended purpose of AI software determines whether it counts as a medical device, so this must be clarified before entering the Australian market.insuranceDiagnostic lab claim denials jump 40% in 2025, XiFin report findshospital managerRising lab claim denials call for attention in revenue management and in discussions with insurers.health ITBilling and dispute-resolution systems need to be ready for a higher denial rate.toolElsevier launches clinical AI tool for nurses as Nextech picks agentic AI partnershealth ITAgentic AI partnerships are moving the market in patient communication automation.toolVitalize raises $31M for AI-powered hospital staffing and scheduling systemhospital managerAI support for scheduling can be a tool for managing staff shortages.health ITIntegrating scheduling systems with control systems is a task to be handled.drug developmentVirtual biotech staffed by 37,000 AI agents finds marker for more successful medicinespharmaceuticalsThe agent-based research model points to a new methodological direction in candidate selection.caseApplications open for sixth cohort of NHS Fellowship in Clinical AIdoctorThe fellowship gives NHS clinicians a clinical AI training opportunity.hospital managerThe programme can be a way to build in-house staff who understand AI.researchDisclosing AI use lowers trust in doctors, especially among younger patientsdoctorDoctors should think carefully about how they disclose AI use, as trust may fall and advice may be followed less.hospital managerPatient information on AI use requires a communication protocol.researchAI took better medical histories than residents in a randomised trialdoctorParts of history-taking can be handed to AI, which affects training and residents' work.

← Issue 9 · Issue 11 →

Issue 10

Utah lets AI prescribe acne medication without direct physician oversight

Who is affected and how

UK government accepts all 44 recommendations of national commission on AI in healthcare regulation

FDA outlines 2027 AI guidance plans; nurses can comment on generative AI framework until October 19

Senator says Medicare AI prior authorization pilot delays care; documents point to a rushed launch

California law says medicine must be practised by doctors and nurses, not AI

Hungary

More key stories of the week

AI companies run virtual drug trials to improve success of human studies

Isomorphic Labs reportedly seeking funding at $40–50 billion valuation

Prospective feasibility study of AMIE conversational diagnostic AI published in The Lancet

AI speeds up skin cancer checks in the NHS, with shorter waits in Sussex and Bradford

AI chatbots often fail users in mental health crises, and ChatGPT for Teens is among those criticised

Three-quarters of FDA-authorized AI medical devices are in radiology; real-world study finds more brain aneurysms detected

Blue Cross report says AI raises healthcare costs; hospitals dispute it

Tech and AI firms lobby federal health officials in CMS-run Slack channel as Medicare promotes commercial health apps

Analysis: AI-enabled drug discovery pipeline is growing but remains largely early-stage

Missing chatbot records leave doctors unable to trace harm from AI health advice

FDA clears first AI test predicting breast cancer risk from a pathology slide; another risk tool available nationwide

Teladoc adds AI scribe for hospitals as Suki launches UpToDate-based support

In brief