What often appears to be turf wars between healthcare technology management, facilities OT staff, IT departments and security teams are often the result of unclear ownership and accountability for device security. And that presents safety risks to patients, says Mohammed Waqas, CTO of Armis.

AI chatbots rank as the number-one health tech hazard in 2026, followed by "digital darkness" and legacy medical device cyber issues, said Rob Schluth and Scott Luney, technology and security experts at patient safety group ECRI Institute, which compiles an annual top 10 list. They explained why.

AI-powered ransomware compresses attacks from weeks to minutes. Michael Villar, director of field security technology at Akamai, says banks need AI-driven segmentation to contain intruders fast, limit lateral movement and protect sensitive data before extortion begins.

Attackers that want to use artificial intelligence tools to build ransomware or help run their cyber operations risk getting much less than they bargained for, said security expert Candid Wuest, in part because they'll still rely on known tactics that can be readily spotted and blocked.

As the compliance deadline quickly approaches for changes to align the federal rules for the confidentiality of substance use disorder records with HIPAA, entities that participate in so-called Part 2 programs still face critical unanswered questions, said attorney Aleksandra Vold of BakerHostetler.

Artificial intelligence use in healthcare is only as safe and accurate as the governance and trust frameworks surrounding it, particularly in clinical environments where errors or hallucinations can directly impact patient care, said Dave Bailey, vice president at consultancy Clearwater.

Britain will likely ban at least some types of ransom payments as it revamps the nation's cybersecurity laws, but many open questions remain, including sectors and the organizational sizes to be covered, and if all payments might be required to pass sanctions checks, said policy expert Jen Ellis.

Revisions to 42 CFR Part 2 that go into effect soon to better align federal regulations for the confidentiality of substance use disorder records with HIPAA require entities to adjust their compliance programs. But the changes aren't easy, said attorney David Holtzman, founder of HITprivacy LLC.

Large language models have a well-earned reputation for making things up. But for AI cybersecurity architect Erica Burgess, rather than being a bug, GPT hallucinations can be a threat-modeling feature. "I like to think of the hallucinations as just ideas that haven't been tested yet," she said.

Use cases of AI in healthcare will continue to expand in 2026 - including for back-office automation, ambient exam room documentation, claims processing and clinical decision support - but so will critical privacy, security, legal and other risks, said attorney Wendell Bartnick of law firm Reed Smith.