See if you can spot the deep lies of AI in our test

If you find that difficult, you are not alone.

It was much easier to spot the computer-generated visual creations — often used by fraudsters — because the AI ​​would make mistakes like adding an extra finger or other obvious oddity.

But AI learns from its mistakes.

“Training on visual artefacts such as looking for a sixth finger or odd earrings has had limited success, as AI has become so good that fraudsters avoid using images with obvious flaws anyway,” said Professor Amy Dawell.

In the picture, Sutherland is holding a woman with shoulder-length hair. The picture of the man is fake.

Dowell is director of the Australian National University’s Emotion and Phys Lab.

She is leading a team of researchers in Australia, Canada and the UK to find out if humans can be trained to beat AI intruders.

The answer, for now at least, is yes — but learning to spot AI liars requires a more subtle approach.