OPINION - This election is a maze of confusing policies, but here's how AI could help

 (Stefan Rousseau/PA Wire)
(Stefan Rousseau/PA Wire)

When will you make up your mind about who to vote for on July 4? Once you’ve properly read the manifestos, right? No doubt you’ve been spending your evenings fully digesting those political page-turners? I didn’t think so…

Research shows that in fact only about 10 per cent of us read them. Instead, we usually vote on a mixture of gut instinct, tribal loyalty and, in some cases, a single issue of particular personal importance.

Most of us have only a hazy sense of what the next prime minister says they will do in office, what pledges they will honour and promises they will break. We haven’t got the time or inclination to learn more and let’s face it, we assume most policy promises are likely to be reversed or binned anyway.

What we desperately need is access to a tool that can guide us through this moral and technical maze of political choices.

Increasing levels of misinformation are clearly problematic but we’re in danger of throwing the AI baby out with the bathwater

Well, that tool exists. Right now. It’s the election guru you didn’t know you had, the machine that will make you smarter, better informed and educated. I am, of course, talking about AI.

Not that you would know that from the headlines that warn only of a wave of misinformation and disinformation impacting the upcoming election here. And we’re not alone. In a landmark year for democracy around the world, machine learning is global enemy No1.

AI badly needs some PR. There are real issues with this still-infant technology that need confronting. Increasing levels of misinformation are clearly problematic but we’re in danger of throwing the AI baby out with the bathwater.

The Alan Turing Institute has revealed there is no actual evidence of AI significantly changing results in any of the 112 national elections which have taken place since January 2023 or are upcoming this year. Out of these 112 elections, only 19 were identified as showing any AI interference at all.

In fact, with the right datasets I believe AI will significantly affect the way we engage with politics.

It can protect us from misinformation by fact-checking, verifying the claims and counter claims filling our news feeds hour by hour. In fact, the technology already exists to answer questions personally important to our needs on the manifesto commitments and show how they compare to each other on any subject, be it childcare costs, housing or transport.

Let’s embrace the magic of the machines and this opportunity for education. And maybe it will help us keep our political leaders honest and make better informed choices than by closing our eyes and imagining who best suits the steps of No 10.

Sarah Golding is Partner of T&Pm