FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Project managers
AI can read and score tenders quickly, but a PM running an evaluation must be able to explain every score and show the process was fair. Losing bidders challenge decisions, and an unexplained AI score is hard to defend.
Part of: AI in project management
In short
Project managers may use AI to summarise tender returns, check compliance with requirements, compare prices and propose quality scores. The risks are scores that do not follow the published criteria, hidden bias against smaller or unfamiliar bidders, and an evaluation that cannot be explained when a bidder asks why it lost. Where individuals are scored, UK GDPR Articles 22A to 22D apply to solely automated significant decisions. Clients may sue the PM if a flawed process leads to a successful challenge or a poor appointment. These claims usually fall under PM PI, subject to the wording.
Common uses include extracting key data from tender returns, checking each bid against mandatory requirements, normalising prices, drafting moderation notes and proposing quality scores against the criteria. On large frameworks, AI can triage hundreds of responses. Some tools also screen suppliers using external data on financial standing or past performance.
The attraction is consistency and speed. The risk is that consistency is only apparent: a model may weight words and length rather than substance, favour bidders whose language resembles past winners, or penalise a bid that answers a question in an unexpected section.
The PM must run the evaluation in line with the client's published criteria and any procurement rules that apply to the client, and must be able to explain each decision. Where the evaluation scores individuals, such as named key personnel or sole traders, UK GDPR applies, and since 5 February 2026 solely automated significant decisions are governed by Articles 22A to 22D, inserted by section 80 of the Data (Use and Access) Act 2025, with safeguards including information, human intervention and a way to contest. For RICS-regulated PMs, the RICS AI standard requires written reliability decisions (4.2) and explanation of the system on request (4.4). On construction projects, appointment competence checks under Part 2A regulation 11E remain the client's duty, often delivered through the PM.
Illustrative scenario. A PM uses AI to score quality submissions for a public sector client. The tool marks down a bidder whose method statement was in an appendix, scoring it as missing. The bidder asks for its scores and an explanation, and the PM cannot show why the tool gave the marks it did. The client has to rerun the evaluation, delaying the project, and seeks its costs from the PM.
A different failure is an appointment made on AI screening that misses a supplier's financial weakness or poor safety record, which then causes delay. Either way, the client's question is whether its PM ran a competent process.
Confidentiality is a further risk. Tender returns contain bidders' pricing and commercial information, often supplied on terms that limit how it is used. Uploading them to a tool that retains or trains on the data may breach those terms and the client's own obligations.
Claims by the client for a flawed evaluation are usually framed as negligence by the PM, which PM PI is designed to cover, subject to the terms. Claims by losing bidders are normally brought against the contracting authority, not the PM, though the client may then seek recovery. Discrimination or data protection allegations can raise policy exclusions or fall to other covers; see AI bias and discrimination claims and GDPR cover within PI.
Apex arranges PI for project managers running procurement, and a broker can check how your policy treats AI-assisted evaluation. See AI resource planning, client data and PM AI tools and the AI in project management hub.
It is risky. Scores must follow the published criteria and be explainable to bidders. A safer approach is to use AI to check compliance and consistency while human evaluators read the bids and own the scores.
They apply where a significant decision about an individual is based solely on automated processing, for example scoring a sole trader or named personnel. Scoring companies is mainly a procurement fairness issue, but data protection still applies to personal data in bids.
Challenges are usually brought against the client running the procurement. The client may then claim against its PM if the PM's evaluation was at fault, which is where PI is commonly engaged.
Transparency is generally sensible and may be expected by the client's procurement rules. At minimum the client should agree the approach, and the evaluation must remain explainable whatever tools are used.
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.
