clinical AI

A new scoping review published by Nature examines how often researchers share the computer code behind clinical prediction models, using artificial intelligence tools to help screen and analyse the literature. The study highlights transparency practices in medical AI research, a topic of growing concern given how rarely model code is publicly available for independent verification.

The American Dental Association is putting a focus on artificial intelligence and technology, drawing attention to how new digital tools are shaping dental practice. The topic is circulating among professionals and observers tracking how AI is being adopted across healthcare fields, including dentistry, and what it could mean for patient care and clinical workflows.

Big Tech companies have committed roughly $700 billion to artificial intelligence, and analysts at Fortune argue that healthcare will determine whether those massive investments deliver returns. The sector is seen as one of the largest potential markets for AI, from drug discovery to clinical documentation, but adoption has been slow due to regulation and trust concerns. Commenters are debating whether medicine can absorb AI fast enough to justify the spending.

Coverage in oncology media highlights how artificial intelligence is being framed as a transformative force in precision oncology, shifting how clinicians match patients to targeted therapies. The discussion centres on AI's role in analysing genomic and clinical data to personalise cancer treatment decisions.

A new clinical trial indicates that AI chatbots could play a useful role in urgent care settings. The findings, reported by the New York Times, add to ongoing debate about whether artificial intelligence tools can safely support medical triage and patient guidance, and how they might be integrated into healthcare systems.

German tech news outlet heise reports on a rivalry between two health AI companies: Caery uses artificial intelligence to audit billing at health insurance funds, while Calliora searches hospital billing data for revenue opportunities. The standoff pits insurers trying to reduce payouts against clinics trying to maximise them, with algorithms on both sides of Germany's statutory health system.

Healthcare startups Ours Privacy, Arintra and Happy Health have secured new venture funding, according to Fierce Healthcare's 2026 funding tracker. The roundup highlights continued investor interest in health technology, spanning privacy tools, AI-powered clinical operations and mental health care. The companies' funding amounts and lead investors were not detailed, but the tracker signals early momentum in digital health dealmaking for the year.

Researchers writing in Nature present symbolic regression as a method for building machine learning models in healthcare that doctors can actually understand. The introduction explains how symbolic regression derives explicit mathematical formulas from data, unlike black-box models, potentially easing adoption in clinical settings where transparency and accountability in medical decision-making are essential requirements.

Healthcare organisations are being urged to evaluate their own readiness before selecting a large language model for clinical use. Writing in Clinical Leader, the argument is that picking the right medical AI depends less on the technology itself and more on whether an organisation has the data, workflows and governance in place to deploy it safely. This kind of readiness-first guidance is drawing attention as hospitals weigh which AI tools to adopt for documentation and decision support.

A clinical trial indicates that AI chatbots could play a role in urgent care settings, according to a report by The New York Times. The finding points to growing evidence that conversational AI tools may support medical triage or patient guidance when immediate care is needed. Details of the trial's size, methods and results were not provided, but the report adds to ongoing debate about how safely and effectively AI can be integrated into frontline healthcare.

A debate is unfolding over whether physicians can preserve independent medical judgment as artificial intelligence becomes embedded in clinical practice. Commentators in the medical community argue that reliance on algorithmic recommendations risks eroding diagnostic reasoning and professional autonomy, while others see AI as a support tool. The discussion reflects growing concern about how much decision-making should be delegated to automated systems in patient care.

A member of a small research group describes a six-month study into whether hospital clinical decision-support workflows could still be trusted if the underlying AI models went dark for four hours. The account suggests the answer exposed heavy dependence on automated tools in medical decision-making, raising questions about fallback procedures, staff preparedness and patient safety when such systems fail unexpectedly.

Researchers at Beth Israel Deaconess Medical Center have carried out what they describe as the first real-world study examining the safety and quality of AI tools used directly by patients in primary care. The research looks at how such systems perform in everyday clinical settings rather than controlled trials. The announcement is drawing attention as health systems weigh expanding patient-facing AI amid questions about accuracy and patient safety.

A guidance piece from Urology Times outlines the key questions urologists should ask when evaluating artificial intelligence tools for prostate cancer care. It highlights the need for clinicians to assess validation, data quality and clinical fit before adopting AI in diagnosis or treatment planning, amid growing use of AI in oncology.

Researchers have published the first real-world study examining the safety and quality of AI tools used directly by patients in primary care settings. The findings are being reported by Medical Xpress and offer an early look at how such systems perform outside controlled trials, a question of growing importance as AI assistants become part of routine healthcare.

A new analysis argues that AI analysis of retinal images could serve as a biomarker to screen and stratify pregnant patients in clinical trials. The piece suggests eye-based data may help identify risk profiles in pregnancy research, where trial enrollment has long been limited by safety concerns. The claim is drawing attention in the clinical research community, though it remains a forward-looking proposal rather than an established practice.

The Agency for Healthcare Research and Quality (AHRQ) has announced a new initiative to use artificial intelligence in medical evidence reviews and data analysis. The American Hospital Association reported the move, which signals the agency's effort to speed up how clinical evidence is synthesized and made available to healthcare providers and researchers.

AI companies are running simulated, or 'virtual', drug trials before moving treatments into human studies, Reuters reports. The approach uses AI models to predict how drugs will perform, with the goal of raising the historically low success rate of clinical trials in humans. If validated, virtual trials could cut the cost and time of drug development, though regulators and researchers will scrutinise how reliable such simulations really are.

A new commentary argues that artificial intelligence in healthcare will not replace doctors' clinical expertise, but instead highlights just how essential that human judgment remains in medicine. The piece suggests AI tools work best when paired with experienced clinicians, reigniting debate about the balance between automation and professional medical knowledge in patient care.

A new analysis in Applied Clinical Trials examines the challenge of deploying artificial intelligence consistently across fragmented oncology health systems. It outlines what hospitals, researchers and technology providers must align on — data standards, workflows and regulation — to make AI tools work uniformly in cancer care across differently organized health systems.

Researchers at Mount Sinai report that adding safety prompts to AI models can guide them toward safer clinical recommendations. The study suggests that simple instruction-based safeguards, rather than costly retraining, may reduce risky medical advice from large language models. The findings are drawing attention as hospitals weigh how to deploy AI tools in patient care settings.

A patchwork of state-level rules governing artificial intelligence in healthcare could delay patients' access to AI-powered tools, according to MobiHealthNews. Without a unified national framework, developers may face differing compliance requirements across states, slowing rollout of diagnostic and clinical AI technologies. The report adds to ongoing debate in the US over how AI in medicine should be regulated and who sets the standards.

Artificial intelligence companies are running simulated, or virtual, drug trials before launching real human studies, aiming to identify promising compounds and reduce the high failure rate of clinical trials. The approach, reported by Reuters, could cut costs and speed up drug development, though questions remain about how reliably computer models can predict outcomes in actual patients.

Researchers at UC Davis Health have launched a clinical trial of an AI-assisted medication device for older adults experiencing memory loss. The tool is designed to help patients manage their prescriptions more reliably and independently. Interest centres on whether AI support can reduce missed doses and ease the burden on caregivers of people with cognitive decline.

Discussion is growing around artificial intelligence being positioned as a collaborative tool for oncologists rather than a replacement for clinical judgment. The framing emphasizes AI supporting cancer specialists in areas like diagnostics, treatment planning and decision-making, with experts arguing its value lies in partnership with physicians. The conversation reflects broader debate in medicine over how AI should be integrated into cancer care while clinicians retain final responsibility.