Letting Agents Report Rise in AI Complaints
On August 11, UK letting agents reported a rise in AI-generated tenant complaints, which Propertymark members described as longer, more legally complex, and sometimes inaccurate. Mortgage Solutions reported that agents are using AI to summarize and analyze lengthy correspondence while retaining human review for responses. Greg Tsuman of ARLA Propertymark also called for complainants to disclose AI assistance under The Property Ombudsman's rules.
UK letting agents are reporting a rise in tenant complaints drafted with generative AI, with Propertymark members describing correspondence that is longer, more complex, and sometimes based on inaccurate legal information. Mortgage Solutions reported on August 11 that complaints which once arrived as calls or short messages are increasingly submitted as lengthy, multipart documents.
Greg Tsuman, a past president of ARLA Propertymark, told Mortgage Solutions that some complaints are without merit and that AI can reinforce a user's assumptions. "AI tends to be very much user-biased," Tsuman said, adding that it can create "a potential false sense of awareness" when a complaint lacks a sound basis.
Legal claims add review overhead
Ben Stokes, ARLA's president-elect and a lettings director, told Mortgage Solutions that tenants have combined online material about previous proposed legislation, including the Renters' Reform Bill, with information that is no longer current. He said agents need to check complaints carefully when the legal position is incorrect.
Property118 reported similar comments from Tsuman and Stokes, including Tsuman's observation that AI tools can produce well-worded complaints far faster than tenants could previously prepare them. The reports do not provide a quantitative measure of the increase, nor do they identify a specific model or AI service used by complainants.
Tsuman proposed that The Property Ombudsman's rules require complainants to disclose whether AI was used to construct correspondence. According to Property118, he said AI assistance should not itself prevent a complaint from proceeding, because it can also make communications clearer for recipients to review.
Agents use AI for triage, not replies
Mortgage Solutions reported that Tsuman's organization has begun using AI to summarize and analyze the substance of lengthy correspondence. Tsuman said the system is not used to generate responses; human staff review the material and respond.
That distinction reflects a broader pattern in high-stakes text workflows: organizations often use language models for extraction, summarization, and document triage while reserving factual assessment, legal interpretation, and external communications for human reviewers. In disputes involving statutory rights, a polished AI-generated document can obscure weak citations or outdated policy references, making provenance checks and claim-by-claim review more important than fluency alone.
The reports also underline a practical evaluation problem for teams building complaint-management or document-intelligence systems. Summarization quality is useful, but systems handling legal or regulatory assertions also need reliable source retrieval, date-aware materials, clear uncertainty handling, and an auditable human review path. Neither retrieved report describes a formal accuracy benchmark or automated legal-validation process used by the agents.
Key Points
- 1Mortgage Solutions reports AI-assisted tenant complaints are longer and more legally complex, increasing review workload while sometimes relying on inaccurate legislative information.
- 2Some letting agents use AI to summarize incoming correspondence while retaining human review and response, according to past ARLA president Greg Tsuman.
- 3Across high-stakes text workflows, fluent generated documents increase the value of provenance checks, current source retrieval, and auditable human review.
Scoring Rationale
The story offers a concrete example of generative AI creating accuracy and review risks in a regulated, consumer-facing workflow. Its direct relevance is strongest for teams deploying document triage, complaint management, or legal-information systems, rather than the broader ML ecosystem.
Sources
Public references used for this report.
Practice with real Real Estate data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Real Estate problems
