Archived on 🗄️

An option to "change expression" on a picture of a person, the option has an AI "glow" indicator next to it.

Reason for archiving

AI is increasingly becoming part of how software is built so will be present in everything, all the time.

Description

Make it clear whether an AI system is in use to create all of, or part of, the interface or experience so that a person can make the right decisions for them. This pattern is important where the information or interface given has a material impact on the person’s choices. This is context specific.

Besides stating it plainly, other common indicators an AI system is in use include a “glow” or a sparkle icon.

IF thinks this pattern will become more important as organisations use more AI systems, and uncover the tacit user needs of making different choices depending on whether an AI is present or not.

Advantages

  • Makes people aware if an AI system is being used or not, and they can make choices appropriate to them.
  • In some contexts, such as debt, research shows that people feel more comfortable when interacting with an AI because they fear less judgement than if they were talking to a person.

Limitations

  • Research shows that in some cases people are more trusting towards AI, not less, and this can lead to people sharing information that can be used against them or stolen by malicious actors.
  • In some cases peoples’ mental model of AI is that it is more accurate than a human. This can create over confidence in an AI.

Examples

  • Spotify's AI DJ →

    When using DJ an icon appears in the app that alerts the presence of the AI setting the track list.