Red Lines

what should never be done with ai and why

Some uses of AI may sit outside what your newsroom is willing to do at all. This card is about naming those limits in advance and the reasons behind them. A team pauses here to agree on what stays off the table before a hard case forces a rushed answer.

Questions to explore

// use these as prompts in a workshop or on your own. There are no right answers.

What uses of AI should your newsroom never allow, and why?
Where would using AI cross a line for your audience's trust?
Which tasks must always stay with a human, no matter how capable a tool becomes?
How do you decide where a firm limit belongs versus a case-by-case judgment?
What would make you reconsider a red line you have drawn?

Expert voices

// notes from the journalists and AI experts who helped shape this kit

“Identify what AI should never do in your newsroom. AI has no sense of consequences, so editorial oversight still matters.”

Martin Schori, Aftonbladet

“Is it acceptable to generate human faces with AI? What risks do you see in generating faces and illustrating real events with AI imagery?”

Anastasiia Korinovska, Suspilne

“Define what must never be automated and where you draw the line. Facial recognition of citizens, for example, should remain off-limits.”

Walid Al-Saqaf, Doha Institute for Graduate Studies

“What is a task you would not trust AI to work on? Answering that honestly is where your red lines begin.”

Carl Javier, Data and AI Ethicist

Things to consider

Limits set in calm moments hold better than limits set under pressure.
A red line is only useful if the whole team knows it exists.
The reason behind a limit matters as much as the limit itself.

using this card

Pull Red Lines when it is relevant and set it aside when it is not. Pair it with the other AI Conversations cards, lay them out on a table, and use the questions above to get everyone on the same page. Capture what you discuss on sticky notes or in a shared doc.