FCA authorised · FRN 724952 0117 325 0027Quote & buy →
AI insurance
Affirmative AI cover is insurance that expressly covers named AI activities or AI-caused loss, while silent AI means a policy neither covers nor excludes AI in clear terms. It matters because insurers are now adding AI exclusions and clarifications, so a policy that was silent last year may not stay silent at renewal.
Part of: AI insurance UK
In short
Most UK business policies were written before generative AI and say nothing about it. That silence cuts both ways: an AI-related claim might be covered because nothing excludes it, or disputed because the insurer never priced it. The market met the same problem with cyber, where the PRA pressed insurers from 2017 to reduce non-affirmative exposure and Lloyd's then required every policy in its market to state whether cyber was covered or excluded. AI is following that path through new exclusions, endorsements and AI-specific policies. Affirmative AI cover reduces the guesswork by naming your AI work in the policy.
The insurance market borrowed these terms from cyber. The Prudential Regulation Authority's supervisory statement SS4/17 (July 2017) defined affirmative cyber risk as policies that "explicitly include coverage for cyber risk" and non-affirmative, or silent, cyber risk as policies that "do not explicitly include or exclude coverage for cyber risk". The market now uses the same terms for AI.
| Term | What the wording does | What it usually means for an AI claim |
|---|---|---|
| Affirmative AI cover | Names AI activities or AI-caused loss as covered, often in the schedule or an endorsement | The insurer has assessed the exposure; the claim is judged on liability and the policy terms |
| Silent AI | Says nothing either way about AI | Cover turns on general words such as the business description, leaving more room for dispute |
| AI exclusion | Removes claims arising from the use, development or output of AI | No cover for that exposure unless another policy or a write-back provides it |
| AI clarification or sub-limit | Confirms AI is covered but caps it or adds conditions | Cover exists, possibly at a lower limit or with conditions such as human review |
Most commercial wordings in use today were drafted before generative AI was in everyday business use. The insurer's worry is less that AI claims exist and more that it may be paying for exposure it never assessed or charged for:
The response is already visible. In the US, ISO, which publishes standard policy forms, released three optional generative AI exclusion endorsements for general liability effective January 2026 (CG 40 47, CG 40 48 and CG 35 08), which can exclude bodily injury, property damage and personal and advertising injury arising out of generative AI. In late 2025 the Financial Times reported, as summarised in legal commentary, that several large insurers had sought US regulatory approval for broad AI exclusions, some reaching any use, or even alleged use, of AI. For UK businesses the point is simple: a policy that was silent last year may come back at renewal with an AI exclusion or clarification. See AI exclusions and clauses for the common forms.
Silent cyber was the market's last large clean-up of this kind, and it shows where silent AI is likely to head.
| Date | Step | What it did |
|---|---|---|
| July 2017 | PRA SS4/17 | Defined affirmative and non-affirmative cyber and expected insurers to "introduce measures that reduce the unintended exposure", such as premium changes, robust exclusions or specific limits. |
| 30 January 2019 | PRA letter to chief executives | Found the largest non-affirmative exposure in casualty, financial, motor and accident and health lines, and that claims teams struggled to identify such claims. Expected an action plan by H1 2019. |
| 4 July 2019 | Lloyd's Market Bulletin Y5258 | Required all policies to be clear on cyber, by exclusion or affirmative cover, starting with property damage lines from 1 January 2020. |
| 29 January 2020 | Lloyd's Market Bulletin Y5277 | Set later phases from 1 July 2020, 1 January 2021 (including professional indemnity, D&O and general liability) and 1 July 2021. |
Two lessons carry over. Clarity did not mean more cover: Lloyd's let underwriters comply by excluding cyber or by covering it affirmatively, and AI clarifications can go either way. And silent exposure is hard to spot at claim stage, as the PRA found with cyber, especially when AI is one link in a chain of events.
One difference: silent cyber was pushed by the PRA and a phased Lloyd's timetable. For AI, change so far has shown up mainly in individual wordings, so read your own policy rather than assume the market has moved in step. For the cyber background, see silent cyber and affirmative cover and silent cyber exclusions in PI policies.
Wordings vary, but affirmative AI cover typically has:
Affirmative cover is not unlimited cover. It can still carry sub-limits, conditions and exclusions, for example for deliberate acts or liability taken on by contract. The value is a clear starting point: your AI work is part of the insured activity. A growing number of policies in the market now give affirmative AI cover in some form.
| A client says your AI forecasting tool gave figures it relied on and lost money (illustrative) | Silent policy | Affirmative AI policy |
|---|---|---|
| Is the activity insured? | Depends whether "software development" is read to include building and selling an AI tool | The named activity matches what you did |
| Was the risk fairly presented? | Open to challenge if the AI work was not disclosed | Disclosed as part of the insured activity |
| Where the argument sits | Coverage, then liability | Liability and the policy terms |
Apex has an AI insurance scheme with an insurer that writes your specific AI work into the policy as insured activity, rather than leaving it to a generic business description. Cover options are professional indemnity, cyber, product and public liability, and IP infringement, so several AI exposures can sit in one policy and you choose the options you need. Cover for US exposure is available, subject to the insurer's underwriting and the policy terms. To quote, we need a precise description of your AI work, your address, your turnover (including how much comes from the USA and Canada) and the limit of indemnity you want, and you can start an AI insurance quote online in a few minutes. For an overview, see AI insurance UK or the short reference page what is AI insurance.
See also: What is silent AI? · What affects the cost of AI insurance
Silent AI is when a policy neither expressly covers nor expressly excludes losses involving artificial intelligence. Cover for an AI-related claim then depends on general wording such as the business description and other exclusions. The term mirrors silent cyber, which the PRA defined in 2017 as policies that do not explicitly include or exclude cyber risk.
It depends on the wording. Many policies say nothing about AI, so cover turns on whether your business description includes your AI work and whether any exclusion applies. Some insurers now add AI exclusions or clarifications at renewal. Read the description, exclusions and any new endorsements, and tell your broker how you use AI.
Affirmative AI cover expressly includes AI activities or AI-caused loss in the policy, usually by naming your AI work as insured activity and dealing with AI output in the insuring clause. It does not remove every exclusion or limit, but your AI work is clearly part of what the insurer has agreed to cover, subject to the policy terms.
It is a close parallel. With cyber, the PRA asked insurers from 2017 to reduce unintended exposure, and Lloyd's then required policies to state clearly whether cyber was covered or excluded, phased in between 2020 and 2021. AI wordings are now going through a similar clarification, so far mainly through new exclusions, endorsements and AI-specific policies.
If your AI use is material to the risk, yes. The Insurance Act 2015 requires a fair presentation of the risk before the policy starts, including every material circumstance you know or ought to know. Building or selling AI, or using it to produce client work, is likely to be material. Leaving it out can reduce or remove cover.
Tell us what your AI work is, where you are based, your turnover and the limit you need. It takes a few minutes online, and a broker will come back to you with options. Or call 0117 325 0027.
Start an AI insurance quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information, not advice on your individual circumstances, and it does not guarantee that cover will be available or on what terms.
