Aipedia: Patent and trade mark attorneys
Attorneys' PI usually covers negligence claims arising from AI-assisted work in the same way as any other negligence, subject to the policy terms. The detail depends on IPReg's minimum terms, any top-up wording and what you told insurers.
Part of: AI in patent and trade mark attorneys
In short
IPReg requires attorneys serving the public to hold PII appropriate to the risks of their practice that meets its Minimum Terms and Conditions. The 2021-22 terms added affirmative cover for claims from a Cyber Act or Cyber Incident (cl.1.1.7), with exclusions at cl.6.5 and 6.6. IPReg lists participating insurers and runs a PII sandbox. No AI-specific statement from IPReg, CIPA or CITMA was found, so AI is usually "silent" in attorney PI. Exclusions or questions may appear in non-participating or top-up wordings. Describe your AI use accurately at renewal.
IPReg's PII requirements apply to registered and licensed firms and sole traders, including attorneys working through firms not regulated by IPReg. Cover must be appropriate to the risks of the practice and meet IPReg's Minimum Terms and Conditions, under Code of Conduct paragraphs 3.9 to 3.12. IPReg publishes a list of participating insurers and operates a PII sandbox that lets regulated persons test cover from non-participating insurers. The Compensation Arrangements Rules 2021 protect consumers and small businesses for fraud or failure to account where PII does not respond.
The minimum terms are written around civil liability for professional work, not around the tools used. An error made using AI is still an error in the professional service. See IPReg PI requirements and attorney PI cover explained.
The 2021-22 Minimum Terms and Conditions introduced express affirmative cover for claims arising from a Cyber Act or Cyber Incident (cl.1.1.7), with related exclusions at cl.6.5 and 6.6. IPReg said these changes preserve the scope of cover. Most AI errors are not cyber events: a hallucinated authority or a misread deadline is a professional error, not a malicious act or system failure. The cyber wording becomes relevant when an AI tool is compromised or leaks data. See silent cyber and affirmative cover.
"Silent AI" means a policy that neither grants nor excludes AI-related loss in terms. No IPReg, CIPA or CITMA statement on AI and PII was found, and no evidence was found of AI exclusions in participating insurers' wordings. In that position, a negligence claim is usually assessed on whether the work met the standard of a reasonably competent attorney, whatever tool was used.
That may not hold everywhere. Top-up and excess layers above the minimum, and policies from insurers outside the participating list, may carry different terms, including AI questions or exclusions. Read every layer. See affirmative versus silent AI cover and AI exclusions and clauses.
Illustrative scenario: a firm adopts an AI docketing system and tells nobody outside the firm. A migration error drops a group of renewal dates and several patents lapse. The clients claim. The primary layer responds, subject to its terms. The claim is large enough to reach the top-up layer, whose insurer asks why the change in deadline management was not mentioned at renewal, given a proposal question about material changes to systems and procedures. The firm now has to argue about fair presentation as well as defend the claim. See AI docketing and deadlines.
Attorney PI is usually written on a claims-made basis, so the policy in force when a claim is made responds, not the one in force when the AI-assisted work was done. Drafting, translation and prosecution errors can take years to surface, often only in opposition or litigation. If you discover that an AI tool produced a flawed output on past files, the policy may require you to notify a circumstance promptly, even before any client complains. Late notification can put cover at risk. Keep records of which files used which tools, so that the extent of any problem can be traced. See claims-made versus occurrence.
The duty of fair presentation applies to all of this. See does PI cover AI mistakes and circumstance notification.
Related: IPReg AI guidance explained, confidential invention disclosures and the attorneys hub.
How Apex helps: Apex arranges IPReg-compliant PI and top-up cover for patent and trade mark attorneys. A broker reviews every layer for AI wording and helps you present your AI use clearly.
No AI-specific statement from IPReg was found, and no evidence of AI exclusions in participating insurers' wordings. Negligence in AI-assisted work is usually treated like any other negligence, subject to the policy terms. Top-up layers and non-participating wordings can differ, so check each one.
Usually not directly. Clause 1.1.7 gives affirmative cover for claims arising from a Cyber Act or Cyber Incident. Most AI errors, such as invented authorities or wrong dates, are professional errors rather than cyber events, and are assessed under the main negligence cover.
If the proposal asks, answer accurately. Even if it does not, a significant change such as AI docketing or AI drafting may be material under the duty of fair presentation. Describing your controls usually helps insurers understand the risk.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI 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 about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.