AI content detection: ‘The days of spotting it by six fingers are over’

TL;DR: Professor David Hughes of Teesside University says visual giveaways in AI-generated content are disappearing fast. He advises checking sources and emotional reactions rather than looking for image flaws. The north-east is meanwhile expanding its AI footprint with growth zones in Blyth and North Tyneside.

As AI-generated images and text become harder to distinguish from human-created content, a government adviser on AI has offered practical guidance to the public. Professor David Hughes, associate dean in research and innovation at Teesside University, says the old tricks no longer work.

Beyond six fingers

“Increasingly, the days of ‘it’s got six fingers’ are going,” Hughes said. Instead, he recommends questioning where content comes from and what reaction it is designed to provoke.

“If it’s trying to spark a big reaction in you, just trying to get you to click on it and engage in it, then that’s never healthy,” he said. His simplest test: “How does this make me feel?”

Hughes pointed people towards trusted, accountable sources and noted that social media platforms have responsibilities to take down misinformation.

AI growth in the north-east

The region is expanding its AI presence with growth zones at sites in Blyth, Northumberland and Cobalt Park in North Tyneside. In Cumbria, the nuclear industry is using AI-trained robots for tasks unsafe for humans. Nissan’s Sunderland plant is using AI to improve car safety.

Energy concerns

On the environmental question, Hughes acknowledged that data processing uses significant energy. “Data uses energy, there’s no question about it,” he said, adding that the industry is “always trying to use green energy” but that responsible use remains important.

Looking forward

As AI-generated content improves, detection will shift from spotting visual flaws to assessing context and intent. Media literacy — not just technology — will be the main line of defence.