Subscribe to Holyrood updates

Newsletter sign-up


Follow us

Scotland’s fortnightly political & current affairs magazine


Subscribe to Holyrood
by Josephine Jay
14 June 2023
Programmed to serve

AI robot Ai-Da in the House of Lords last year | Alamy

Programmed to serve

As we career through the twenty-first century with all its technological advancements, so too come with it anxieties over human obsolescence in the face of robotic innovation. While some might be thrilled at the prospect, there are considerations to be taken in before we scrap elected politicians for mechanical alternatives. 

The obvious advantage of a robotic politician lies in their perceived moral imperviousness. For example, a robot would not fail to declare HMRC investigations, it would not become sticky-fingered with party donations because the foibles attached to robotic nature are not those imbued in the human condition. A robot is mechanically honest to a fault; a robot would not have an affair with an aide, it would not drive to test its eyesight. A robot would remember the number of children it had. 

Within this point lies a wider one about the lack of trust in politicians held by the general public. Our preference for robotic politicians lies in our belief in the moral turpitude of our human ones. Recent headlines have fixated on the private lives of those elected to lead our country with the idea that the quality of one’s politics is driven by personal morality - an indicator certainly but by no means a deciding swing factor. Worldwide trust in politicians is at an all-time low. Reports across the board show cynicism with a 2021 UK poll reporting two out of three people felt politicians were merely ‘out for themselves.’ Furthermore, a study in 2021 showed that 51 per cent of Europeans would happily replace lawmakers with artificial intelligence (AI) with the term ‘algocracy’ (government by algorithm) creeping into vernacular. 

AI features in politics whether we like it or not. Israeli president Issac Herzog delivered a speech at a cybersecurity convention in Tel Aviv earlier this year part-written by AI, while US congressman Jake Auchincloss, the Massachusetts Democratic Representative became the first politician to deliver an AI-generated speech on the floor of the House. If harnessed correctly, AI has the potential to expedite and elevate political decision-making. If exploited, AI has the potential to turn the ugly beast of politics even uglier.  

While AI is not distracted from policy-making by sex drive, shirking taxes or self-interest, it is still under the sway of political bias. A study conducted by the Technical University of Munich and the University of Hamburg showed that ChatGPT – an AI chatbot released in 2022 with over 100 million monthly active users in its first two months – was influenced by political bias and censorship. It showed ChatGPT had a “pro-environmental, left-libertarian orientation” meaning its stances on issues such as undocumented migration, abortion rights and taxing the rich were decidedly left-leaning. It refused to propagate incendiary material, as shown by a Forbes article in February, where ChatGPT declined to pen poems about former US president Donald Trump presumably as it felt this violated its hate-speech coding. Additionally, the 2020 Netflix documentary ‘Coded Bias’ showed clear racial bias within AI facial recognition technology, demonstrating how this reinforces data bias, perpetuates inequality and leaves minority groups vulnerable to statistical targeting. 

AI by its nature is open to misuse. The geopolitics lurking behind the large companies that fund, operate and monitor AI are even further removed from the public eye than some MP tax returns. AI has the potential to be weaponized – previous elections and referendums have proven only too well the sway misinformation and scare tactics can hold. AI’s main strengths lie in its ability to process large amounts of data incredibly quickly. It can trawl social media absorbing public sentiment, ascertaining trends and responding to these as they emerge. Therefore its ability to mass generate content has the capability to fuel political agendas under the guise of media and by burying politics in fluff. AI is also fallible. Just as students find ever-inventive methods of side-stepping anti-cheating software - so too would people manipulate and circumvent AI politics. 

We need politicians to remain human just as we need humanity to remain at the forefront of politics. Many of the pressing issues of our time, healthcare, migration and climate change are issues demanding compassion. Policy-based solely on numbers, dots and dashes is a cold and binary process that would ultimately fail those who need it most. The answer lies in collaboration and adaptation; AI has the potential to enhance aspects of political processes, however, the responsibility of integration ought to be one laced in transparency, accountability and safeguarding designed to enhance democracy not exploit it. 

An AI chatbot’s answer to whether it thought it would make a good politician was predictably bland; ‘I am not programmed to have political ambitions or affiliations.’ It concluded, ‘I could potentially make a fair and neutral politician…my purpose is to assist.’ The point is, would you vote for that?

Josephine Jay is a participant in the Pass the Mic project which aims to improve the representation of women of colour in the media. Holyrood magazine is very happy to once again be involved with Pass the Mic and is supportive of all its ambitions.

Holyrood Newsletters

Holyrood provides comprehensive coverage of Scottish politics, offering award-winning reporting and analysis: Subscribe

Get award-winning journalism delivered straight to your inbox

Get award-winning journalism delivered straight to your inbox


Popular reads
Back to top