Opinion

Ai, ai for AI faking it…

It used to be that any villain, unmasked on the pages of a newspaper for something he or she said, would cry “I was misquoted” or “I was quoted out of context”. But a colleague recently pointed me to the new, and much more difficult to disprove, excuse: “the video (or soundclip) was faked.”

It used to be comparatively simple to determine whether a piece of electronic communication had been doctored. Artificial Intelligence (AI) makes that trickier now.

This week, on X, I watched a spoof video put together by creative people using easily available AI tools.

Advertisement

It showed all the world’s leaders in various odd situations: Kamala Harris drunkenly swigging on a bottle of wine; Donald Trump puffing a cigar in a limo surrounded by hookers, Vladimir Putin in tweeds, brandishing a hunting rifle as Ukrainian flags flutter behind him; Hillary Clinton running with ghetto gangs, bottle of vodka in one hand and a .45 automatic in the other…

The amazing thing about the video was how real it was – I believed I really was looking at a video of French president Emmanuel Macron pedalling a kiddie’s tricycle while holding a baguette.

ALSO READ: ‘Nobody wins in a war’

Advertisement

And therein lies the new refuge for scoundrels. AI will make plausible, believable, content and the internet will allow it to spread at many times the speed of the Covid virus.

We’ve already seen the damage that a little disinformation sprinkled on a bucket of suspicion and scepticism will cause, as millions of Americans (and other conspiracy theorists around the world) are hoping anti-vaxxer Robert F Kennedy jnr will ban vaccines completely when he takes over his key health portfolio in Trump’s incoming administration.

Ironically, I was reminded this week that, even before the internet extended the reach of propaganda infinitely, governments and their spooks were already experts at producing and disseminating disinformation.

Advertisement

A production company putting together a film on the South African Airways Helderberg tragedy asked me for some of my recollections.

The plane went down in the Indian Ocean on 27 November, 1987, after a disastrous onboard fire which many believed was caused by ammunition or explosives being smuggled into South Africa.

ALSO READ: There’s more to China than meets the eye

Advertisement

I believe there was definitely some substance to that… I personally spoke to a CCB (Civil Co-operation Bureau) operative who had brought “sheet explosive” (which looks like A4 paper but can blow off a hand, or a head) into the country aboard an SAA plane from Europe.

It wasn’t long, though, before the professional disinformation agents from Military Intelligence got to work on some of the journalists covering the story… and more and more outrageous “facts” began to adorn the front pages – from explosive concealed in fish tanks to the destructive, but phantom, substance called “red mercury.”

When we realised the absurd direction the coverage was taking – and where the information was coming from – my colleague Peta Thornycroft and I bailed out.

Advertisement

Helderberg stories, though, became so ridiculous that they led to a book claiming the rightwing in South Africa had captured nuclear weapons and a bomber to deploy them.

The result was that no serious journalist wanted to touch the Helderberg story for 20 years – and many still run from it.

ALSO READ: Far from the madding crowd

It was one of the most effective disinformation operations in history, yet their only AI was Apartheid Intelligence…

For more news your way

Download our app and read this and other great stories on the move. Available for Android and iOS.

Published by
By Brendan Seery
Read more on these topics: Artificial Intelligence (AI)crashOpinion