As the UK government expands its use of AI, a new study shows voters want MPs to stay in charge of political decisions
Voters are open to artificial intelligence helping MPs, but reject machines making political decisions, reveals research from Brunel University of London. Asked if there's place in Parliament for AI, people in the UK and Japan are clear that the use of support tools are fine but decision-making by AI bots is not.
The study, published in Parliamentary Affairs, surveyed 990 people in the UK and 2,117 in Japan. People were asked whether AI should help parliamentarians make better decisions, and whether AI or robots should make decisions instead of MPs.
It comes as the UK government widens its use of AI across departments publishing its Blueprint for Modern Digital Government, in January. Speaking at OpenAI's Frontiers conference in London Deputy Prime Minister David Lammy said AI could boost trust in government and cut bureaucracy. Earlier, Conservative MP Tom Tugendhat warned MPs not to outsource their speeches to ChatGPT, after spotting Americanisms.
Eighty per cent of UK study respondents said they oppose letting AI or robots make political decisions instead of MPs. Japanese people were slightly more open to automation, researchers found, but still wary of handing over political authority.
"Politicians may be tempted to lean on AI," said Steve Pickering, Honorary Professor at Brunel Business School, Brunel University of London. "But our research shows voters draw a clear boundary. They may accept AI to sift evidence or stress-test policy, but they want transparency, accountability and above all, humans to call the shots."
Voters are not anti-technology, researchers found. Many are open to MPs using AI to sift evidence, analyse data or test policies. But there is a clear red line when it comes to political judgment. People want human representatives to remain responsible, even if machines work in the background.
Support for AI in politics varies by age, gender and trust. Younger men were more open to its use, while older people and women were more sceptical. Trust in government and confidence in using technology were strong signs of support. The single strongest predictor of opposition was fear. People who said things like 'AI scares me' were much less likely to back its use in politics.
Political thinking shaped attitudes in the UK but not in Japan. In the UK, people on the political right were more supportive of MPs using AI. In Japan, support was consistent across the political spectrum, suggesting ideology played little role.
Legitimacy is what matters most, the researchers say. Voters care about how decisions are made, not just the outcomes. Parliamentary democracy depends on visible accountability and a clear human chain of responsibility. AI may help speed up government, but if public consent is lost, trust goes with it. "AI may be fast, but democracy depends on consent," said Prof Pickering.
-Ends-
Notes to editors
The paper, "Democracy by algorithm? Public attitudes towards AI in parliamentary decision-making in the UK and Japan" by Dr Steve Pickering and Dr Martin Hansen of Brunel University of London and Dr Yosuke Sunahara at Kobe University is open access: https://doi.org/10.1093/pa/gsaf050.
Distributed by Pressat