Responsible use of generative AI in writing
October 22nd, 2026 at 11:36 am (Artificial Intelligence, Writing)
Writing is hard (but also rewarding). How should we guide ourselves (and our students) in when, where, and how to make use of (or not) generative AI systems in writing?
The best answer is context-dependent. Not all writing has the same purpose (or requirements, or implications). Is it a press release? A letter to your grandmother? A scientific research paper? A course assignment? A helpdesk question? A helpdesk answer? A legal document?
That said, we can identify some general principles. Sophie Alpert, the engineering lead at Clay, created an AI writing policy that I think is sensible, supportive, and important. Note how each of the four key points is followed by a reason why it makes sense. It’s worth reading her full policy (it’s not that long), but here are my takeaways:
- Take responsibility for the final product.
- Make sure you’re still engaging your brain.
- Spend more time creating the document than it takes to read it (respect your readers’ time).
- Shorter is better. If your idea is better communicated by sharing the prompt, just do that.
I think this can be a great starting point for conversations about appropriate use of generative AI within a given context.