First published in The Integrity Gap on 2026-09-17. Read the original on Substack.
During the Fund Shack podcast recording (wherever you listen to podcasts here, and the full episode on YouTube), we discussed the need for brevity and better questions amongst the noise.
If you don’t have time for the full recording just yet, in this YouTube Short I described the issue I see becoming increasingly common. A larger party (in this case, an investor, but it could be a corporation) asks the counterparty (prospective investor, third-party, supplier, etc.) to complete a form.
From what I am seeing, the problem is that DD forms are becoming unthinkingly burdensome and irrelevant. The larger party might argue, “I don’t have time to customise DD forms for every counterparty.” My retort would be: anything that large is (fodder for) AI slop, and maybe you should learn to use AI to shrink, and not expand, the dataset.
We’ve been finessing exactly that this week: a document review tool that builds on a contextual risk right-size follow-up questions. We call this suite of tools Ranulph, after my beloved dog who passed in 2023 (and because the etymology of his name: advice/counsel/power and wolf).
Slop in, slop back
If we don’t triage down the questions, the smaller counterparty, who (surprise, surprise) also has Claude, will just take the 500-question dirge and get AI to help complete it, or worse… develop a policy suite to make it look like everything requested has been implemented.
The larger party will then not read that massive questionnaire response (and accompanying disclosure), but instead go to Claude (or a DD provider using LLMs) to ask: “Do these answers and policies address our questions?”
The LLM, with no schema, ontology, framework, or any other contextual governance, will then “hallucinate its balls/t!ts off [a delightful tech term I learned this week].” And the larger party will proceed with unearned confidence, as the smaller party goes back to doing what they’ve always done.
This has to stop; it’s nuts.
For the love of brevity
Putting my money where my mouth is (in this case, spending time to create a free resource), I put together the “Beyond the Surface” tool on the Insights page of our website (freely downloadable, no email or sign-up). It’s a practical tool that includes questions to assess a potential counterparty’s understanding of and management of integrity risk.
What you will get: 10 questions on integrity risk, and a few bonus political and security risk questions.
If it’s useful, let me know. In the interim, if you’re wondering why AI (even with a clever prompt) still isn’t helping, get in touch. You need data governance, then software, not prompts. And, yes, that software can be integrated into your existing ways of working.