
Trust and AI in Parliament: Who makes the decisions?
People distinguish between AI supporting politicians and AI taking decisions for them. We compare attitudes in the UK and Japan.
Parliamentary Affairs
TrustTracker / Research themes
Acceptance of AI is about more than whether a system works. We examine how trust, confidence, fear and expectations of human responsibility shape support for using it.

Respondents in both countries were more supportive of AI helping parliamentarians than of AI taking decisions for politicians.
Our studies examine how trust in government, scientists and other people relates to acceptance of AI.
Confidence in using AI, optimism about its benefits and fear of the technology help explain who is receptive and who is sceptical.
Average support for two uses of AI in politics, on a five-point scale.
November 2024 surveys. Analytic samples: UK 990; Japan 2,117. Scale: 1 = no support; 5 = complete support. Values are descriptive means from the published table.
Source: Pickering, Hansen & Sunahara (2025), Parliamentary Affairs, Table 1 →
| Use of AI | UK | Japan |
|---|---|---|
| Helping parliamentarians | 2.12 | 2.48 |
| Taking decisions for politicians | 1.4 | 2.21 |
For public institutions, introducing AI is also a question of legitimacy: what role do citizens expect humans to retain? Our research distinguishes support for a tool from support for handing over a decision.

People distinguish between AI supporting politicians and AI taking decisions for them. We compare attitudes in the UK and Japan.
Parliamentary Affairs

Trust in government, scientists and other people helps explain public trust in AI, alongside optimism, fear and confidence in using the technology.
AI & Society

Who supports AI in public administration? We compare how trust and individual characteristics shape attitudes in the UK and Japan.
International Journal of Public Administration

Willingness to share health records depends on the recipient and the wording used. Anonymisation alone does not necessarily reassure people.
British Journal of Healthcare Management (31)
Our ongoing work on personal-information disclosure asks how people navigate the boundaries between useful technology, institutional trust and control over their data.
Which policy packages make the use of personal information more acceptable? Experimental research compares public preferences in the UK and Japan.
Eric Margolis draws on TrustTracker research to examine the gap between Japan’s AI ambitions and workers’ experiences.
We explain where public opinion draws the line between AI assisting politicians and taking decisions for them.
Read about the survey design and research team, or access the data and replication code.
About the research → Open data & code →The original TrustTracker research programme was supported by the Japan Society for the Promotion of Science (JSPS, grant reference JPJSJRP 20211704) and the UK Research and Innovation's Economic and Social Research Council (UKRI-ESRC, grant reference ES/W011913/1).