Where AI shouldn't go in your business

The short answer

Keep AI away from work that depends on judgement, accountability or relationships: hiring and firing, pricing strategy, difficult client conversations and anything where a person has to stand behind the decision. AI can help prepare for that work. It shouldn't do it.

Judgement

AI produces plausible answers, which isn't the same as good judgement. When a decision weighs things that aren't written down anywhere, like a client's history, a team's morale or the risk you're willing to take, it needs a person who understands the context.

Accountability

If something goes wrong, someone has to explain why the decision was made. "The AI suggested it" isn't an explanation a client, employee or regulator will accept. Hiring decisions are the clearest example: they affect people's livelihoods, and they need a person who can account for them.

Relationships

Clients and staff can tell when a message was written to them and when it was generated for them. Use AI to draft routine replies. Write the difficult ones yourself.

Use AI to prepare for the decision. Make the decision yourself.

Where it still helps

• Summarising the background before a hard conversation.
• Drafting interview questions, then deciding with people.
• Pulling together the numbers before a pricing decision.

Part of every readiness audit is a clear list of what to leave alone. Knowing where not to use AI is as useful as knowing where to use it.

Written by Daine Mawer

Founder of Plaincraft, an independent AI consultancy for South African SMEs, with 17 years of building websites and digital products behind it. AI is part of how I work every day, so I know where it earns its keep and where it doesn't.

More about me