On 6 October 2026, the Commission for Communications Regulation (“ComReg”) published guidance (the “Guidance”) on the use of Artificial Intelligence (“AI”) in formal end-user dispute resolution processes led by ComReg. The Guidance sets out ComReg’s expectations of providers who leverage AI in these dealings with ComReg.
Scope
ComReg is entrusted with a statutory role in formal dispute resolution of disputes between end-users and providers of certain telecoms and postal service providers, and the Guidance is focused on this particular context. However, the Guidance is likely to provide a good indication of ComReg’s expectations in relation to the use of AI by regulated companies more generally, including in the context of regulatory compliance investigations and dealings with end-users in general.
The Guidance highlights the known risks of using AI, including that it may produce inaccurate content, and that personal and confidential data may be improperly processed, and sets out the below core requirements to mitigate against those risks.
Core requirements
The Guidance outlines four expectations relating to AI use in end-user dispute resolution:
- Responsibility for AI-generated content: Parties remain fully accountable for all submissions to ComReg, whether made directly or on their behalf, and ComReg treats all submissions as originating from and approved by the relying party. Failure to take responsibility for AI-generated or AI-assisted content may undermine the credibility of a party’s case, require information to be corrected or withdrawn, and result in process delays, thereby detracting from the overall effectiveness of the dispute resolution process.
- Accuracy and completeness of AI-assisted submissions: Parties must ensure all information relied upon is accurate, complete, relevant, and reliable, and, where available, provide supporting evidence and documentation to substantiate this. ComReg reserves the right to seek clarifications or require information to be resubmitted if it receives a submission from a party that is excessively long or complex, or forms the view that a submission has been produced with the assistance of AI.
- Verification requirements for AI-generated information: Parties must independently verify the accuracy, completeness, relevance and reliability of any AI-generated content against authoritative sources before submitting it to ComReg. Parties must not use the same or any another AI tool to verify such content. Failure to appropriately verify AI-generated content may require clarification or correction, delay the dispute resolution process or weaken the reliability or credibility of a submission.
- Disclosure obligations for AI use: Parties must acknowledge AI use in their submissions and be prepared to explain the nature and extent to which AI was utilised. ComReg may seek confirmation from parties regarding the accuracy, completeness, relevance and reliability of any submission where it suspects AI assistance. Disclosure does not remove, transfer or reduce responsibility. Even where AI use is disclosed, the submitting party remains responsible for ensuring the accuracy, completeness, relevance and reliability of any submissions.
Key takeaways
While ComReg welcomes the responsible use of AI to streamline submissions and other communications, the Guidance mandates explicit safeguards to target the recognised risks of AI.
The following checklist may assist service providers when using AI in ComReg-related matters:
- Has the fact of AI use been properly disclosed?
- Has there been sufficient human oversight of any AI-generated materials?
- Has any information generated by AI been verified?
Contact us
If you have any questions on this article, please contact Matheson Partners Kate McKenna, Simon Shinkwin or your usual Matheson
