With a general election imminent, we need to think seriously about the effects of AI and digital manipulation in our own country. Democracy is a fragile apparatus. To operate well, it requires accurate information.

Before he was actually indicted, AI-generated fake images of Donald Trump scuffling with the NYPD during his “arrest” swiftly accumulated millions of views. Proof again that Artificial Intelligence (and the internet) exudes a unique magic to entertain, dissimulate and also deceive.

Despite proclamations that the startling pictures were false by Eliot Higgins, the man who—with the help of an AI called Midjourney—created and shared them, the images were often received (and disseminated) not as jokes, but as truth.

“In an unprecedented turn of events,” wrote one blogger and ex-U.S. infantryman on Facebook, “former President Donald Trump was arrested and escorted to federal prison.” The Washington Post reports that major tech platforms such as Twitter and Facebook failed to always indicate that the pictures were crooked.

Ah, you might think, such is America, the polarised post-truth political jungle where information (and disinformation) is instantly weaponised. How sad. Luckily we’re not like that in our own dear part of the world.

Actually, no. These are issues with which any technological society must contend.

In last year’s Australian election, an internet strategy helped so-called Teal (fiscally conservative, i.e. blue, but environmentally concerned, i.e. green, mixed, therefore teal) independent candidates. Targeted not with outright lies but messages intended to sway voters; the Teals tipped Liberal MPs out of supposedly safe seats. Digital experts used YouTube, and particularly Facebook for messaging. Oh, and money. In one case, a Teal candidate outspent her Liberal enemy by four to one. She triumphed.

A recent survey of trust in the news shows that it has dropped alarmingly (by about one-sixth) in two years.

Digital nudging is a spectrum; what’s that antidote? A strong fourth estate might help. Only, we don’t have one of those here. A recent survey of trust in the news shows that it has dropped alarmingly (by about one-sixth) in two years. The risk profile increases. New Zealanders trust businesses more than they do the government. Eroding faith in democratic institutions creates a vacuum, something nature and information (good or bad) abhor.

So with a general election imminent, we need to think seriously about the effects of AI and digital manipulation in our own country. Democracy is a fragile apparatus. To operate well, it requires accurate information.

Legislation might be one answer. Unfortunately, good laws are generally slow laws, and tech is much faster. The race to legislate the internet has left lawmakers in the US and Europe scratching their heads.

Any other solutions? Perhaps it’s you, dear reader. New Zealanders are highly distrustful of social media. Double down on that. Scrutinise at every image or link you see on your feeds. Does the person depicted doing that crazy thing have, say, three legs, like the Trump “arrest” images (I’m serious) did? Where did the video, image or link come from? If it smells rotten, it likely is. Move on.

An antipodean version of the Trump deep fakes featuring say Jacinda Ardern, Chris Hipkins or Chris Luxon is inevitable. Don’t let it sway your vote or others’. Just lol, and keep scrolling.