This article contains a reference to suicide.

This is The Guardian.

The Guardian Archive Long Read.

Hi, I'm Amit Katwala. I'm an Editor at Wired magazine and the author of The Guardian Long Read,

the Race to Create a Perfect Lie Detector and the Dangers of Succeeding, which was published in 2019.

I first got interested in the subject while I was watching Making a Murderer,

which was a Netflix show, a true crime show about a guy called Stephen Avery who had been,

he says, wrongly accused of a couple of crimes that he said he didn't commit.

That case got worldwide attention in the Netflix series Making a Murderer.

Every season of that show, there was a test that he did to try and clear his name called a brain

fingerprinting test, which purported to be able to tell if someone was lying based on a brain scan.

My background is in experimental psychology and I was immediately quite skeptical of this technology

and I started researching it, I interviewed the creator and that sent me down the rabbit hole of

the new lie detection technologies that I look at in the piece. So brain scans, AI, voice analysis

and all these other things that I'd try and talk about in the article.

The article traces the long history of lie detection from ancient China through to the present day,

but particularly the polygraph machine and then the new things that have sprung up to try and replace

the polygraph machine. The underlying premise of the piece basically says that these technologies

can never really work and that we are simply repeating the same mistakes that were made a century

ago with a polygraph machine with these new types of AI lie detector and that despite

some concerns over how they work, they are still being snapped up by governments and police

forces and companies that want to stamp out lying.

...