Healthcare goes operational AI

Hospitals and drugmakers are pushing artificial intelligence deeper into day-to-day operations, from missed-appointment calls to early-stage drug hunting. (prnewswire.com) Luma Health said on April 21 that its Spring 2026 release adds “Operational AI” tools that automatically call patients after no-shows, reschedule visits in real time, read incoming faxes for clinical findings, and trigger follow-up workflows without staff handoffs. The company said its earlier AI workflows have already saved 2.5 million hours and handled more than 350,000 care-related next steps. (prnewswire.com) Luma said the new package includes a no-code Workflow Builder, a Workflow Gallery based on customer requests, Navigator for phone-based no-show recovery, and Fax Transform for extracting findings from documents such as mammograms and colonoscopies. The company said it serves more than 1,000 healthcare organizations and 100 million-plus patients across the United States, United Kingdom and Caribbean. (prnewswire.com) Drug discovery uses software to sift through biological data and chemical structures before a medicine ever reaches human testing. Insilico Medicine said on March 29 that Eli Lilly will use its Pharma.AI platform in a collaboration spanning multiple therapeutic areas and preclinical oral drug programs. (prnewswire.com) Insilico said the deal gives Lilly an exclusive worldwide license on the covered programs, pays Insilico $115 million upfront, and could reach about $2.75 billion in development, regulatory and commercial milestones, plus tiered royalties on future sales. Lilly’s Andrew Adams said the partnership is aimed at finding promising candidates faster across several disease areas. (prnewswire.com) The same push to automate work is colliding with data-access risks. ShareGate said on April 21 that 29% of organizations in a March 2026 survey reported AI tools had already surfaced sensitive data they should not have accessed. (prnewswire.com) In that survey of more than 850 information technology and security leaders in the United States, Canada and Europe, 36% cited exposed customer records, 31% sensitive internal documents, 30% personal data, 30% human resources records, 25% financial data and 21% intellectual property. Only 51% said they had completed an organization-wide governance review after enabling Microsoft 365 artificial intelligence tools. (prnewswire.com) In healthcare, those governance questions sit on top of older privacy rules. The U.S. Department of Health and Human Services says its Office for Civil Rights has initiated 2024-2025 Health Insurance Portability and Accountability Act audits focused on Security Rule provisions most relevant to hacking and ransomware attacks. (hhs.gov) That leaves providers and life-sciences companies trying to prove who approved an AI workflow, what data it touched, and what evidence supports each action. As more healthcare work moves from staff queues into software, the paperwork around permissions and audit trails is moving with it. (prnewswire.com)