It’s in your search engine. Your word processor. Your browser. Your grammar checker. Your research platform. You didn’t choose to work in an AI-enabled environment – opening your laptop put you in it.

So when a contract, partner, client, or associate says, “no AI,” what does that mean?

It’s not about workflow – you might list all the tools you use, but that’s going to be pretty much every app, software, or browser extension you work with. It’s more about the deliverable. Human reviewed has been accepted as a reassurance, but that’s not the same as human created. Pretty much most practice I’m seeing sits somewhere in between.

So where is the line?

Source discovery feels defensible to me. I’ve had RSS aggregators for years, and Perplexity is a more informed version of that – identifying candidate documents, then opening every source manually, checking it, and deciding whether it’s relevant. That’s research support where Perplexity returned a lead list, and I did the work.

I know you’re fed up with reading more about AI, but it’s everywhere, and it’s important we have a conversation about it.

What isn’t defensible? That’s using AI to draft, rewrite, edit, refine or proofread and then not disclosing it. If you’re being commissioned for professional guidance, it’s about your expertise and judgement, not just words.

The ethical standard has to be disclosure. That means agreeing on boundaries, logs, and their auditable use. Which tool was used, when it was used, what was it used for, and how was it integrated into the work? It’s not about ‘research’; it’s about the queries you ran, the prompts you used, and the instructions and parameters you put in place to prevent the tool from overreaching.

This is what we should be talking about with commissioners, contributors, and vendors. Saying “I didn’t use AI” won’t cut it any more; you did use it.

We need to understand just how (and where) if we’re paying for the work.

---

#AIEthics LearningAndDevelopment #ProfessionalDevelopment #AIgovernance