Your Business Insurance May Have Dropped AI Coverage in January 2026

Advertising & Editorial Disclosure

Picture a five-person marketing shop that uses an AI writing tool to draft client ad copy. A campaign goes out, a competitor claims the copy infringed on its trademarked slogan, and a lawsuit follows. The shop owner calls her broker, expecting her commercial general liability policy to cover legal defense costs for certain advertising-injury claims. Instead, the insurer points to an endorsement added at her last renewal: an exclusion for claims connected to generative AI.

That kind of denial became easier for insurers to write after Verisk made new exclusion endorsements available to carriers in January 2026. Many small business owners are unaware that their policies may now include these optional AI exclusions.

On Jan. 1, 2026, Verisk's ISO program introduced three optional generative AI exclusion endorsements for commercial general liability policies: CG 40 47, CG 40 48 and CG 35 08. The forms aren't automatic, and adoption varies by carrier, state and renewal date. But they raise a new coverage question for any small business using AI in marketing, customer service, product work or everyday office software, whether the liability coverage sits in a standalone CGL policy or a business owner's policy. Brokers and coverage attorneys have started advising clients to check renewal paperwork for the new language.

What Changed in January 2026?

Verisk added optional generative AI exclusion endorsements for commercial general liability policies starting Jan. 1, 2026. ISO, Verisk's forms division, publishes standard policy language widely used by commercial insurers. Before these endorsements, many standard CGL policies didn't directly name generative AI. Coverage analysts called the arrangement "silent AI" coverage, a situation similar to the cyber exposure carriers spent years resolving after 2018. When attached, Verisk's new endorsements remove that silence and state directly that claims tied to generative AI aren't covered under the affected coverage parts.

The change is optional for insurance companies, not automatic for every policyholder. An insurer has to choose to attach CG 40 47, CG 40 48 or CG 35 08 to a specific policy. That choice depends on the state, the insurer's underwriting rules and the policy's renewal date. A business owner whose policy renewed in late 2025 may not see the exclusion yet.

Businesses that lose this coverage do not have an automatic replacement. Technology errors and omissions coverage, cyber insurance and standalone AI liability policies can each respond to some of the same risks. None is a direct substitute for the general liability coverage a business had before.

What CG 40 47, CG 40 48 and CG 35 08 Exclude

CG 40 47 is the broadest of the three forms. It removes Coverage A (bodily injury and property damage) and Coverage B (personal and advertising injury) for any claim "arising out of" generative AI, on occurrence and claims-made policies alike.

CG 40 48 is narrower. It excludes only Coverage B, personal and advertising injury, the part of a CGL policy that responds to claims like defamation, copyright infringement or trademark violations tied to AI-generated marketing content.

CG 35 08 applies to a different coverage part, products and completed operations liability. It excludes bodily injury or property damage connected to generative AI in a product a business made or sold.

CG 40 47
CGL Coverage Part (occurrence and claims-made)
Coverage A (bodily injury, property damage) and Coverage B (personal and advertising injury) arising out of generative AI
Claims not arising out of generative AI under this coverage part
CG 40 48
CGL Coverage Part (occurrence and claims-made)
Coverage B (personal and advertising injury) arising out of generative AI
Coverage A: bodily injury and property damage
CG 35 08
Products/Completed Operations Liability Coverage Part
Bodily injury or property damage arising out of generative AI in a product
Premises and operations exposure

All three forms share the same definition of generative AI: “a machine-based learning system or model that is trained on data with the ability to create content or responses, including but not limited to text, images, audio, video or code,” according to Verisk's filing language. The definition, which Verisk says draws on government and National Association of Insurance Commissioners sources, covers chatbots, AI writing tools, image generators and AI code assistants. It doesn't extend to older, non-generative AI tools that only score or classify data instead of producing new content.

Why “Arising Out of AI” May Reach Everyday Business Tools

In insurance law, "arising out of" is a broader standard than plain English suggests. The phrase requires only a causal connection to the excluded activity, not direct or sole causation, according to Pillsbury attorneys Hadhy Ayaz and Jay Konkel, who wrote in Policyholder Pulse in April 2026. The attorneys cautioned that if courts read the phrase broadly, insurers may argue that a wide range of claims are connected to AI use, even when AI played only a minor role in a loss.

Insurance broker Arthur J. Gallagher & Co. gives a concrete example in a 2026 client alert. An AI-assisted drawing or calculation contains an error that leads to a structural failure or an injury. Even when an employee reviewed and approved the work before it went out, an insurer can still argue the loss “arises out of” generative AI and deny the claim. Gallagher also says the exclusion can apply whether a business used AI directly or picked it up secondhand through a vendor, consultant or software platform.

That secondhand exposure extends the exclusion well beyond businesses that built their own AI tools. A business doesn't need custom software to trigger it. The exclusion can reach any small business using AI features built into software it already pays for, from customer relationship platforms to email and document tools, even when that use never registered as a technology decision.

How Many Small Businesses Use AI Now?

Generative AI use among small businesses reached 58% in 2025, up from 40% in 2024 and 23% in 2023. The figures come from a U.S. Chamber of Commerce Technology Engagement Center survey of thousands of businesses with fewer than 250 employees, published in its 2025 Empowering Small Business report. The same report found 82% of small businesses using AI added workers over the past year. Among all small businesses, 84% plan to increase their use of technology platforms.

Small Business Generative AI Adoption Is Climbing Fast

Share of small businesses reporting generative AI use, by year

Source: U.S. Chamber of Commerce Technology Engagement Center, 2025 Empowering Small Business report.

Note: Measures share of small businesses, not workers.

The Main Street AI Monitor, a survey from the U.S. Chamber of Commerce Foundation conducted with Ipsos and published in June 2026, found half of all workers at small businesses already use AI at work. It also found adoption differs by company size: only 43% of businesses with two to nine employees report using AI for work tasks, compared with 59% of businesses with 100 to 249 employees.

Among small business AI users, 90% use it for writing and editing communications. Research is the next most common use, at 88%, according to the same survey.

These two activities can trigger a Coverage B claim under CG 40 47 or CG 40 48. A business doesn't need to build a chatbot or train a model to have AI exposure. An employee who drafts an ad, summarizes a report or answers a customer email with an AI tool gives an insurer grounds to argue the exclusion applies if that work later leads to a defamation, copyright or trademark claim.

Has Your Policy Changed?

PYMNTS reported that carriers including Berkshire Hathaway, Chubb and Travelers have sought state approval to exclude AI-related claims from general liability policies. State regulators have approved more than 80% of the AI-exclusion filings submitted across the industry. Florida, Connecticut and Maryland approved the highest volume, according to the same reporting. Some approved exclusions began attaching to policies as early as January 2026.

Separately, W.R. Berkley, AIG and Great American Insurance Group sought approval to exclude AI-related liabilities from broader corporate coverage, including directors and officers and errors and omissions policies, as the Financial Times and Barron's reported in November 2025.

Insurers are moving at different speeds. A filing doesn't always mean an exclusion is already active. AIG told regulators it had filed a generative AI exclusion but had no plans to implement it at the time, Barron's and the Financial Times reported.

Gallagher's John Farley told Business Insurance in April 2026 that "there are a few carriers that are starting to adopt those exclusions" and that "we're just at the very beginning." Zurich North America's Kara Higginbotham told the same publication that "many carriers are currently evaluating the potential benefits and risks of making broad policy language changes."

With these endorsements available and some carriers already adopting them, small business owners should assume a policy could carry an AI exclusion rather than assume coverage is unchanged. A business owner can ask an agent or broker directly whether CG 40 47, CG 40 48 or CG 35 08 appears on a current commercial general liability, business owner's policy or products/completed operations form. The form numbers appear on the policy's declarations page, endorsement schedule or policy forms list, and a broker can confirm whether any are present in minutes.

AI Risk May Also Affect Cyber, E&O and Media Coverage

A commercial general liability exclusion isn't the only place AI risk can show up in a business's insurance program. Cyber insurance often responds to security or privacy breaches, but claims over AI hallucinations, defamation or intellectual property infringement from AI-generated content fall outside most cyber policies, according to Financial Times reporting on how insurers are pricing AI risk.

Technology errors and omissions coverage is more likely to respond to AI-related mistakes. Media liability coverage addresses content, defamation and intellectual property claims tied to what a business publishes. A growing number of tech E&O, cyber and media carriers have added their own AI exclusions or carve-outs that track language similar to Verisk's CGL forms.

Marsh's Greg Eskins told Business Insurance in April 2026 that “the market has been restrained in taking any drastic actions with either affirmative or exclusionary language, though both exist.” PYMNTS and Independent Agent reported that a handful of standalone AI liability products have entered the market for businesses that use third-party generative AI tools in customer-facing work, including offerings from Armilla AI, Munich Re, Corgi and Testudo.

What to Ask Before the Next Renewal

Business owners should ask a broker or agent to confirm whether a current general liability, business owner's policy or products/completed operations policy carries CG 40 47, CG 40 48 or CG 35 08. A broker can explain which coverage parts, if any, are affected. Questions worth raising directly include:

  1. 1
    If my policy has an AI exclusion, which claims would fall outside my coverage?
  2. 2
    Does my cyber or technology errors and omissions policy include AI-related exclusions?
  3. 3
    If I use AI in marketing or customer communications, what coverage would respond to a defamation or copyright claim?
  4. 4
    Would a standalone AI liability policy cover any risk my current program leaves open?

Business owners should review a full insurance program, not only the general liability policy. AI exposure can reach cyber, technology errors and omissions, media liability and directors and officers coverage. Any business that uses AI in customer-facing work or embeds it in its products should ask how each policy would respond to an AI-related claim.

Frequently Asked Questions

Common questions from small business owners about the new generative AI exclusions, whether they apply to a current policy and what coverage remains.

About Myryah Irby


Myryah Irby, Writer and Data Journalist

Myryah Irby is a writer and data journalist at MoneyGeek. Her work spans original data studies and how-to guides covering auto, home and health insurance, consumer costs and transportation safety.

Research and Analysis

Since joining MoneyGeek in late 2025, Irby has produced data studies on insurance costs, consumer spending and transportation risk. Her published work includes a 50-state analysis of winter driving danger using fatality and weather severity data; research tracking the relationship between rhodium commodity prices and catalytic converter theft rates, including state-level theft trends and what those rates mean for insurance costs; a state-by-state comparison of winter home heating costs; and an analysis of the full cost of having a baby in America: hospital bills, insurance and out-of-pocket expenses.

Career

Irby has more than 20 years of editorial and writing experience. Since 2005, she has run Irby x Irby, her own editorial and copywriting practice, with clients including The New York Times, The San Francisco Chronicle, OpenAI and the National Park Service. From 2019 to 2023, she served as Senior Managing Editor and then Copywriting Manager at Callisto Media, a nonfiction publisher acquired by Penguin Random House in May 2023, where she led a team of writers and graphic designers.

Before that, she spent nearly 11 years at QuinStreet, a performance marketing company that runs content and comparison sites in insurance and personal finance. She rose from Managing Editor to Senior Managing Editor between 2010 and 2016. Earlier in her career, she edited at Collabrys for nearly four years and tutored doctoral candidates on dissertation writing at the University of San Francisco.


Sources