The honesty question: when to tell customers you used AI
Disclosure is not about guilt. It is about not letting people believe something untrue about what they are getting.
Every business using AI eventually hits the same question. Do we need to tell people? The honest answer is that it depends on what your customer believes they are buying.
A useful test
Ask whether a reasonable customer would feel misled if they found out later. Nobody feels betrayed that a spellchecker fixed a typo. Many people would feel misled if a "personal letter from the founder" was generated in bulk, or if a product photo showed something that does not exist.
Where disclosure matters most
Disclosure carries the most weight when a person is the point of the product. Advice, coaching, art, testimonials and anything framed as one human speaking to another all fall in this group. It also matters for images and audio that could be mistaken for real events or real people.
Where it matters less
Internal drafts, research help, formatting and routine operational messages rarely need a label. A human reviewing and standing behind the final result is what customers care about.
Keep it simple
A short, plain note works better than legal language. Something like "drafted with AI assistance and reviewed by our team" tells people what they need to know without making it a drama.
Some places now have rules about labeling AI-generated content, so check what applies where you operate. Beyond the law, trust is the asset. Being straightforward about how you work is one of the easiest ways to keep it.