CFC has added explicit, affirmative AI coverage wording to its global eHealth product and its US Professions policy, with equivalent endorsements available on Professions policies elsewhere, the London specialist insurer said on October 27.
The eHealth wording addresses AI errors, bias, hallucinations and failures involving tools such as ambient scribes and clinical trial support software, according to Coverager's report of the statement. CFC's full suite of Healthcare package products has also been upgraded to affirmative AI cover, and the Professions update sets out how professional liability responds as firms put AI into client services and daily operations. CFC disclosed no limits, sublimits or pricing for the new wording.
The extension comes four months after CFC embedded affirmative AI language across seven products in June, and six days after Chubb chief executive Evan Greenberg told analysts that AI liability "is not contemplated in casualty underwriting really to date." More than 60 US property and casualty carriers have filed AI exclusions, with more than 80% of the filings approved, according to a CSIS count published on September 4.
Key Takeaways
- The eHealth wording names AI errors, bias, hallucinations and failures, and CFC's full Healthcare package suite was upgraded to affirmative AI cover at the same time, Coverager reported.
- CFC underwrites AI developers to training data and accuracy rates, upfront and ongoing, US healthcare lead Kyle Laudadio said on October 12: "We can't have even 1% go sideways here."
- CFC's October 29 update to its cyber policy added enhanced affirmative AI wording and stretched the business interruption indemnity period from 12 to 18 months.
- Most of the US commercial market is moving the other way: more than 60 carriers have filed CGL AI exclusions, with over 80% approved, per CSIS figures reported in September.
What the eHealth wording covers
CFC introduced its eHealth form in early 2017 for telehealth organizations and has followed the market into AI since, Kyle Laudadio, who leads CFC's US healthcare practice, said on the Professional Liability Underwriting Society's podcast on October 12. "One thing we've done recently is affirmatively updated our language to include artificial intelligence as a coverage trigger within our policy," he said.
The form can sit on either side of a scribe claim, Laudadio said. CFC may insure the technology company that built the AI, the clinician using it, or both, and a documentation error could turn out to be a medical malpractice matter, a cyber breach that altered the record, or a flawed, hallucinating algorithm. A single policy carrying all three coverages avoids "the multi-carrier finger-pointing," he said.
From the clinician's side, he expects such claims to land on medical professional liability. "Ultimately, you're the one who should have been signing off that everything's correct," he said.
On the developer side, CFC underwrites to the training data, Laudadio said: where it came from, whether patient permissions were obtained, how wide the data set is and which frontier model the tool was built on. "We want to see the accuracy rates upfront and ongoing accuracy rates," he said. "We can't have even 1% go sideways here."
The rollout since June
CFC began the program on June 26, when it added explicit AI language to Tech E&O, Professional Liability, eHealth, Intellectual Property, Management Liability, Media and Cyber Proactive Response, naming model hallucination, AI-generated content and model drift as exposures, Reinsurance News reported. "Rather than relying on implied or silent coverage, we see value in being explicit about how AI is treated," Nick Line, CFC's chief underwriting officer, said at the time.
Both eHealth and Professional Liability were on the June list. The October wording names the healthcare tools and perils, extends the Professions change to the US form, and makes the equivalent endorsement available on Professions policies in other countries, according to Coverager.
| Date | Product | What changed |
|---|---|---|
| June 26 | Seven products, Tech E&O to Cyber Proactive Response | Program announced; hallucination, AI-generated content and model drift named |
| September | Financial institutions suite | AI used against the policyholder confirmed within the cyber section |
| September 23 | Intellectual Property | Affirmative wording across training data, algorithms and AI outputs |
| October 27 | eHealth, US Professions, Healthcare package | AI errors, bias, hallucinations and failures; ambient scribes and clinical trial software named |
| October 29 | Cyber Proactive Response | Enhanced affirmative AI wording; BI indemnity period 12 to 18 months |
Sources: Reinsurance News, June 26, September 23 and October 29, 2026; Insurance Business, September 17, 2026, via actuary.info's Beazley report; Coverager, October 27, 2026.
The IP wording, added September 23, confirms that "where an AI-related intellectual property claim would otherwise be covered, the involvement of AI alone does not create uncertainty around how the policy responds," Maddi Brown, CFC's IP practice leader, said in the Reinsurance News report. CFC said the wording clarifies the policy response without altering its scope.
Two days after the healthcare announcement, CFC released an update to its Cyber Proactive Response policy with enhanced affirmative AI wording, personal crime and extortion protections for senior executives, and a business interruption indemnity period extended from 12 to 18 months, Reinsurance News reported on October 29. "Cyber risks continue to evolve rapidly, whether through the growing adoption of AI technologies, increasingly targeted attacks against senior executives or the longer-term consequences of business interruption events," Scott Bailey, CFC's head of global cyber underwriting, said.
The exposure a scribe affirmation takes on
The affirmation runs against the direction of the US commercial market. Verisk's CG 40 47 family of endorsements, effective January 1, 2026, lets carriers remove generative AI from commercial general liability. State regulators had approved more than 80% of the exclusion filings by April 23, according to CSIS, and ISO forms sit under roughly 82% of US P&C policies. An April analysis of the forms put the AI load being removed at 0.5 to 2.0 points of expected loss ratio on a standard CGL policy.
Laudadio described the same split in his own market. "Most policies in the market right now are silent on AI. I'm hearing that at the beginning of the next year, there actually could be exclusions for AI," he said, adding that a healthcare insured would need to know whether an exclusion reaches billing, intake and scribes or only clinical care.
For the carrier affirming the peril, the exposure is correlated and thin on data. An ambient scribe vendor serves many practices on one model, so a transcription error, a bias in the model or a hallucinated entry is one event that can reach many insureds at once. Laudadio said a breach at a third-party AI vendor "could be a systemic event across multiple health systems," with potentially millions of affected individuals.
The June analysis of the first CFC rollout put the reporting tail for hallucination claims at 12 to 36 months and found the trigger date for model drift ambiguous, and the analysis of Chubb's call put full credibility for a new liability class at about 1,082 claims. CFC's underwriting answer, per Laudadio, is the accuracy rates it asks for upfront and on an ongoing basis, which stand in for the claims history that does not yet exist.
"There are some other markets that are following us," Laudadio said. "I suppose that there's going to be more competition in that space in the next couple of months."
Further Reading
- Affirmative AI Coverage Moves Model Drift Into Pricing Territory: the June CFC rollout and the four perils' reporting tails.
- Beazley Writes AI Into Cyber Cover as 60 Carriers Exclude It From CGL: Beazley's attacker-side endorsement and CFC's financial institutions wording.
- Chubb's Greenberg Says Casualty Books Have Not Priced AI Liability: the clarification, exclusion, standalone cover sequence from the Q3 call.
- CGL AI Exclusions Win 80% State Approval as Carriers Shed Generative AI Risk From Standard Books: the approval rate and which carriers filed.
Sources
- Coverager, "CFC extends AI coverage to healthcare and professional services," October 27, 2026
- PLUS Podcast, "Your Ambient Scribe is Listening: Is Your Policy Ready?" with Kyle Laudadio of CFC, October 12, 2026 (transcript)
- Reinsurance News, "CFC introducing affirmative AI coverage," June 26, 2026
- Reinsurance News, "CFC introduces affirmative AI cover for intellectual property risks," September 23, 2026
- Reinsurance News, "CFC launches executive cyber protection for SME leaders and upgraded AI cover," October 29, 2026
- Coverager, "CFC adds affirmative AI coverage across insurance portfolio," June 2026