Skip to main content

TrustTracker / Research themes

AI can assist. But who decides?

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.

Robot legislator

What we’re learning.

01

Assistance and delegation are different

Respondents in both countries were more supportive of AI helping parliamentarians than of AI taking decisions for politicians.

02

Trust carries across domains

Our studies examine how trust in government, scientists and other people relates to acceptance of AI.

03

Experience and expectations matter

Confidence in using AI, optimism about its benefits and fear of the technology help explain who is receptive and who is sceptical.

Helping politicians is more acceptable than replacing them.

Average support for two uses of AI in politics, on a five-point scale.

https://doi.org/10.1093/pa/gsaf050 | 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. image/svg+xml Matplotlib v3.10.8, https://matplotlib.org/ 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Mean support, 1–5 Helping parliamentarians Taking decisions for politicians 2.12 1.40 2.48 2.21 UK Japan Pickering, Hansen & Sunahara (2025), Parliamentary Affairs, Table 1

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 →

Read the chart values
Helping politicians is more acceptable than replacing them. Mean support, 1–5
Use of AIUKJapan
Helping parliamentarians2.122.48
Taking decisions for politicians1.42.21

Why it matters.

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.

The research behind the findings.

All publications →
Robot legislator

UK & Japan · 2025

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

A laptop keyboard, hands and ChatGPT.

UK & Japan · 2026

What makes people trust AI?

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

AI & Society

A small white guidance robot standing in a Japanese public-service building beside a sign explaining that a robot customer guidance experiment is underway

UK & Japan · 2026

Trust, AI and public administration in Japan and the UK

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

Patient consent form

England · 2025

Trust and health data

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)

Where we’re going next.

Our ongoing work on personal-information disclosure asks how people navigate the boundaries between useful technology, institutional trust and control over their data.

Beyond the journal.

All media & impact →

The survey programme and open data.

Read about the survey design and research team, or access the data and replication code.

About the research →   Open data & code →

Research support

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).

JSPS and UKRI/ ESRC logos