But like automation threatening jobs, it seems politicians realise there’s a real need to get to grips with deep fakes. Deep fakes are computer-generated videos and audio, based on real footage of a person. Imagine the impact of a convincing video of Donald Trump declaring nuclear war, except he’d actually done no such thing.
An even bigger threat is from 'evidence’ of things that aren’t easily disproven. Take a video of a politician confessing to a crime in private, for example. Even if it was a complete fabrication, the signal to doubt the trustworthiness of that individual would have been seeded in many people’s minds.
And the potential for fraud on individuals outside public life is notable. Catch someone off guard with a convincing 'phone call’ from a relative asking for emergency cash and they may fall for it.
Now, the US politicians who raised this issue are mainly concerned about deep fakes as a national security issue – foreign powers using the technology to influence Americans. But the domestic threats from this tech (which is likely to soon be good enough and cheap enough for widespread use) are perhaps more serious.
The question is, how do you legislate against this? If you can catch offenders after the fact then you can prosecute them, but it may be impossible to stop deep fakes from permeating society. And we’ll have to get used to never quite knowing for sure how real something is. Our understanding of the world may never be the same again.