Specialist insurer CFC updated its business insurance lineup on June 25, 2026 to embed affirmative AI coverage across seven core products. This means policyholders can now secure explicit protection for risks like model hallucination and AI-generated content, moving past the ambiguity of older, silent policies. For any US business deploying or building AI tools, understanding this shift is now central to evaluating whether your coverage actually matches your exposure.
If your company uses a large language model to draft customer emails, a machine-learning algorithm to screen applicants, or an AI copilot to write code, you have probably asked one uncomfortable question: does our business insurance actually cover this? Until recently, the honest answer for most policyholders was maybe, maybe not. Commercial general liability and cyber policies were written for a pre-generative-AI world, leaving a gap that brokers call “silent AI”, exposure that is neither clearly covered nor clearly excluded. On June 25, 2026, London-based specialist commercial insurance provider CFC took a concrete step to close that gap, launching affirmative AI wording across seven of its core business insurance products. We reviewed the update against industry press coverage and the policy language published by CFC to help US business owners, risk managers, and founders understand what changed, why it matters, and how to evaluate your own coverage in light of it.
CFC’s June 25, 2026 update embeds affirmative AI coverage across seven business insurance lines, at a time when tech firms, SaaS companies, and other AI-deploying businesses face a widening gap between carriers embracing AI coverage and those adopting sweeping AI exclusions.
What “affirmative AI coverage” actually means
In insurance, “affirmative” coverage is the opposite of “silent” coverage. A silent policy says nothing about AI, leaving a carrier and a policyholder to argue after a claim whether a chatbot error is a professional service mistake, a tech product defect, or a cyber event. Affirmative language, by contrast, explicitly states that certain AI-related events are covered, and it defines the terms. CFC’s June 2026 update inserts explicit AI-related language directly into the policy wording of seven products, according to CFC’s announcement. For a US policyholder, this reduces the risk of a coverage dispute exactly when you can least afford one.
What CFC’s June 2026 update covers
CFC embedded the new AI language across seven products: Technology Errors and Omissions, Professional Liability, eHealth, Intellectual Property, Management Liability, Media, and Cyber Proactive Response. The breadth is deliberate. Nick Line, CFC’s Chief Underwriting Officer, said in the company’s press release that AI “acts as an accelerant of existing risk, reinforcing the need to embed AI considerations across our existing products,” rather than a wholly new category, which means it touches nearly every line of coverage. The updated wording explicitly addresses three novel exposures that have worried risk managers: model hallucination, AI-generated content that infringes IP or defames, and model drift where an algorithm’s performance degrades over time in ways that harm a client, as reported by Insurance Journal and Business Insurance.
Why this matters for US businesses right now
CFC is a London-based specialist commercial insurance provider whose policies are distributed through US surplus lines brokers and cover many American tech firms, professional service providers, and mid-sized enterprises. The update signals a market shift that US-domiciled carriers are watching closely. As of mid-2026, some US insurers have moved in the opposite direction, adopting AI-related exclusions to cap accumulation risk, according to industry press coverage. If your current business insurance was bound before 2026, it likely contains neither affirmative AI grants nor the new ISO AI exclusions, leaving you in the silent zone that CFC specifically designed its update to eliminate.
The “silent AI” problem and why insurers are moving now
Silent AI is not a hypothetical concern. When a generative AI tool produces factually wrong advice that a business then passes to a client, the resulting claim does not fit neatly into a 2018-era professional liability form. The insurance industry is currently reassessing AI-related exposure with a significant focus on coverage definitions and contract language rather than pricing, according to reporting from Business Insurance and Reinsurance News. CFC’s move reflects increasing demand from policyholders for transparency and certainty as AI adoption accelerates across all industries. In other words, customers asked carriers to stop being vague, and some insurers have responded.
How the policy handles model hallucination and AI-generated content
Model hallucination, when an AI system confidently generates false information, is one of the hardest risks to insure because it blurs the line between a product defect and a professional error. CFC’s updated Technology E&O and Professional Liability forms address this directly. The coverage extends to liability arising from AI-generated content that a business publishes or delivers to a client, including errors, omissions, or misleading statements the AI produced. For a US digital marketing agency or a SaaS company embedding a third-party LLM, this language provides a clearer path to coverage than a standard tech E&O policy purchased even 18 months ago.
Model drift: the overlooked exposure in your policy
Model drift occurs when a machine-learning model’s outputs degrade because the real-world data it encounters diverges from its training data. An underwriting algorithm that begins approving riskier loans over time, or a predictive maintenance tool that misses equipment failures, are classic examples. CFC’s 2026 wording brings model drift into scope as a covered cause of loss under its Professional Liability and Tech E&O forms. This is notable because most standard business insurance policies are silent on performance degradation that happens gradually rather than through a sudden, identifiable error. Risk managers evaluating their own programs should ask their broker whether gradual algorithmic deterioration would trigger coverage or fall into a gap.
Beyond cyber: why AI coverage belongs in multiple policies
Much of the market discussion around AI insurance has focused narrowly on cyber policies. CFC took a broader approach, integrating coverage across seven products to reflect AI’s wide impact on business risk. This matters for US buyers because purchasing a standalone cyber policy with some AI language may leave your professional liability, media liability, or management liability exposure unaddressed. A discrimination claim arising from an AI hiring tool, for instance, lands in Management Liability (which typically includes employment practices coverage), not cyber. An AI-generated marketing campaign that infringes a competitor’s copyright lands in Media or Intellectual Property. CFC’s multi-line approach is a useful benchmark: when you review your own business insurance, map each AI application your company uses to the specific policy that would respond to a related claim.
| AI Risk | CFC Policy Line That Responds | CFC’s 2026 Approach |
|---|---|---|
| Model hallucination in client deliverables | Technology E&O / Professional Liability | Affirmative grant with defined terms |
| AI-generated content that defames a person or company | Media | Explicitly included in wrongful act definition |
| AI-generated content that infringes a patent or copyright | Intellectual Property | Affirmative coverage under updated wording |
| Model drift causing client financial loss | Professional Liability / Technology E&O | Covered as cause of loss |
| AI in a clinical decision support or telehealth tool | eHealth | AI exposures embedded in health tech wording |
| AI hiring tool produces discriminatory outcome; board challenged on AI oversight | Management Liability | AI-related governance and EPL exposure acknowledged |
| AI system exploited in a cyber attack vector | Cyber Proactive Response | Integrated into incident response coverage |
A market divided: affirmative coverage vs. AI exclusions
Not every insurer is following CFC’s lead. Two new ISO (Insurance Services Office) endorsements, CG 40 47 and CG 40 48, took effect in January 2026 and permit carriers to exclude generative AI claims from standard commercial general liability policies. CG 40 47 is the broader endorsement, excluding generative AI losses under both Coverage A (bodily injury and property damage) and Coverage B (personal and advertising injury). CG 40 48 is narrower, excluding only Coverage B. A third endorsement, CG 35 08, addresses AI-related claims through a limited exception structure. According to reporting from Business Insurance, several US carriers, including Berkshire Hathaway, Chubb, and Travelers, have begun adopting these endorsements at renewal.
A separate trend involves what carriers describe as absolute AI exclusions. Berkley, for instance, has introduced sweeping AI exclusions across its Directors and Officers, Errors and Omissions, and fiduciary liability lines, barring coverage for claims “based upon, arising out of, or attributable to” any actual or alleged use, deployment, or development of artificial intelligence, according to reporting from Policyholder Pulse. That kind of language, layered on top of the ISO CGL endorsements, creates a market where two businesses in the same industry could have radically different AI coverage depending on their carrier and policy vintage. The divergence makes it essential to read your policy’s exclusions section as carefully as its insuring agreements. An exclusion titled “Artificial Intelligence” or “Algorithmic Decision-Making” may appear in a renewal quote without your broker flagging it verbally.
How to evaluate your business insurance for AI gaps
Start by listing every AI tool your business uses or plans to deploy within the next 12 months, including third-party products like CRM copilots, embedded LLM features in project management software, and custom models your data science team builds. For each tool, identify the worst-case scenario: a hallucinated recommendation that loses a client money, generated content that triggers an IP lawsuit, or a biased output that draws regulatory scrutiny. Then ask your broker, in writing, which specific policy and which specific clause would respond to that scenario. If the answer is vague or references “general coverage intent,” you may be sitting in silent AI territory. Request a copy of your carrier’s most recent AI-related endorsements or exclusion language. As of mid-2026, some US carriers have published these; others have not, and that absence is itself a data point.
For core commercial lines coverage, digital-first US carriers such as biBerk, backed by Berkshire Hathaway, and Ergo Next Insurance, backed by Munich Re, offer streamlined online quotes for general liability, business owner’s policies, workers’ compensation, and professional liability. Neither of these carriers currently underwrites standalone affirmative AI coverage in the way CFC’s updated wording does. It is worth noting that Munich Re, Ergo Next’s parent company, does offer standalone AI liability products to enterprise clients through its specialty lines, alongside a small group of specialty AI insurers such as Armilla, Mayflower Specialty, and Embroker. Businesses with meaningful AI deployments will typically need to combine a core commercial insurance program with a specialty AI liability or affirmative AI-enabled policy accessed through a broker experienced in emerging technology risk.
What the CFC update does not solve
The CFC wording is a meaningful step forward, but it has boundaries. First, it is an insurance product, not a compliance shield; it covers liability after something goes wrong, though some endorsements may respond to fines and penalties stemming from AI regulations. Second, affirmative coverage still requires the policyholder to meet conditions like prompt notice of circumstances and reasonable security controls. Third, the policy defines covered AI events using specific technical language; if your AI use case falls outside those definitions, a claim could still be denied. Finally, CFC’s update applies to its own policy forms. A US business insured with a domestic carrier that has not adopted similar language still faces the silent AI gap, regardless of what a London market insurer offers elsewhere.
Bottom line
- CFC’s June 25, 2026 update embeds affirmative AI coverage across seven business insurance lines: Technology E&O, Professional Liability, eHealth, Intellectual Property, Management Liability, Media, and Cyber Proactive Response.
- The update directly addresses model hallucination, AI-generated content, and model drift, eliminating the ambiguity of silent AI, where coverage was neither granted nor excluded.
- Some US carriers are moving in the opposite direction, adopting ISO endorsements CG 40 47 and CG 40 48 (effective January 2026), and in some cases layering absolute AI exclusions across D&O, E&O, and fiduciary lines.
- AI risk spans multiple insurance lines, cyber, professional liability, media, IP, management liability, and specialty health, so a standalone cyber policy is insufficient to address the full exposure.
- Evaluating your own business insurance requires mapping each AI tool to a specific policy and getting written confirmation from your broker on coverage applicability.
- Affirmative AI coverage is an evolving area of contract language; what counts as a covered AI event depends on precise definitions that vary by carrier.
- Insurance protects against third-party claims and certain first-party losses; it does not substitute for AI governance, testing, and compliance programs that reduce the likelihood of a claim in the first place.
Frequently asked questions
What is affirmative AI coverage in business insurance?
Affirmative AI coverage means the policy explicitly states that certain AI-related events are covered and defines key terms such as AI model, AI-generated content, or model drift. It is the opposite of silent coverage, where the policy says nothing about AI, leaving coverage uncertain. CFC’s June 25, 2026 update embeds this language across seven product lines.
How does CFC’s new AI coverage address model hallucination?
CFC’s updated Technology Errors and Omissions and Professional Liability forms explicitly include liability arising from model hallucination, when an AI system generates false or misleading information. The policy defines the term and treats hallucination as a covered cause of loss, removing the ambiguity that existed in older forms.
Which types of business insurance policies are updated by CFC for AI?
CFC embedded affirmative AI coverage across seven products: Technology Errors and Omissions, Professional Liability, eHealth, Intellectual Property, Management Liability, Media, and Cyber Proactive Response. The multi-line approach reflects the insurer’s view that AI accelerates risks across nearly every coverage area.
Why are insurers adding explicit AI language to policies now?
Insurers are responding to accelerating AI adoption across all industries and rising policyholder demand for clarity. The industry is currently focused on coverage definitions and contract language rather than pricing, according to industry press coverage. Explicit language reduces the risk of coverage disputes and helps carriers manage accumulation risk transparently.
What is silent AI in insurance?
Silent AI describes a situation where a business insurance policy neither explicitly covers nor explicitly excludes AI-related losses. This creates uncertainty for both the policyholder and the carrier, often leading to disputes after a claim. CFC’s 2026 update was designed specifically to eliminate silent AI by introducing defined, affirmative terms. Two ISO endorsements, CG 40 47 and CG 40 48, take the opposite approach, allowing carriers to exclude generative AI claims from commercial general liability policies.
How does AI impact existing business risks for insurance purposes?
AI acts as an accelerant of existing risks rather than creating entirely new categories, according to CFC Chief Underwriting Officer Nick Line. For example, AI can amplify professional liability risk if a model gives bad advice, increase media liability exposure through generated content, or heighten employment practices risk through biased algorithmic decisions. This is why CFC integrated AI coverage across seven policy lines rather than siloing it in a cyber product.
Related BestGuide reviews:
Biberk
Next Insurance
The Hartford