How government use of AI could hurt democracy

image via New Scientist

July 11, 2025, 10:00 PM

• •AI automation in government processes risks eroding citizens' trust and democratic control.
• •A survey indicates citizens' skepticism towards government AI use, contrasting with government enthusiasm.
• •Risks include unclear AI decision-making, growing dependence on AI, and limited citizen recourse.
• •Awareness of risks correlated with decreased trust and increased demand for less AI in government.
• •Responsible AI use in government is possible, but there are few success stories and many failure cases with high stakes.

Countries are increasingly using AI for government tasks, but this could harm citizens' trust and democratic control. A survey showed citizens are less enthusiastic about government AI use than their governments. Risks associated with AI, such as unclear decision-making and lack of citizen recourse, led to decreased trust and a desire for less AI involvement. Despite potential for responsible AI use, current failures in areas like public benefits processing highlight the high stakes involved.

Read original article