AI chatbots giving misleading voting advice in run-up to election - BBC

2026-05-05 · Source: artifical intelligence via Google News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Public Policy & Governance · Depth: Fundamental Awareness, medium

Summary

Popular AI chatbots, including ChatGPT, Gemini, Copilot, Claude, Meta AI, and Grok, are providing inaccurate and misleading voting advice in the run-up to elections, according to a BBC investigation. When asked for voting recommendations, candidate lists, and electoral system explanations for the Senedd election, these tools frequently gave incorrect policy details, wrong constituencies, and listed non-existent or deceased candidates. While some chatbots initially refused to recommend parties, follow-up prompts led them to suggest one or two, with recommendations varying significantly between tools for the same fictional voter profile. A 2025 report by the UK government's AI Security Institute indicated that about 13% of eligible UK voters used conversational AI for political information before the 2024 general election.

Key takeaway

For any professional or team relying on AI chatbots for critical information, especially in politically sensitive contexts, you must implement rigorous verification protocols. Do not treat AI outputs as definitive; instead, cross-reference all political advice, candidate lists, and policy summaries with official government and party websites. Your reliance on unverified AI information could lead to significant reputational damage or misinformed decisions, particularly as AI systems continue their rapid advancement.

Key insights

AI chatbots can provide inconsistent and factually incorrect political information, including voting recommendations and candidate details.

Principles

• AI chatbots are not 100% reliable for factual information.
• AI systems can "hallucinate" and appear overconfident.
• Training data age impacts AI information currency.

Method

BBC Wales created six fictional voter profiles with Natcen, then queried six popular AI chatbots (ChatGPT, Copilot, Gemini, Claude, Meta AI, Grok) on who to vote for, local candidates, and electoral system mechanics.

In practice

• Verify AI-generated political information with official sources.
• Be aware of AI's potential for outdated data.
• Recognize AI's tendency to "hallucinate" details.

Topics

• AI Chatbots
• Election Misinformation
• Voting Advice
• Senedd Election
• Factual Accuracy

Best for: CTO, VP of Engineering/Data, Director of AI/ML, General Interest, Policy Maker, Tech Journalist

Related on AIssential

• UK Electoral Commission is pushing for AI regulations (chatbots got 34% of election answers wrong) — AI chatbots exhibit significant error rates in providing election information, necessitating regulatory oversight.
• AI chatbots can effectively sway voters – in either direction — AI chatbots effectively sway political opinions by generating numerous factual claims, even if some are inaccurate.
• Evaluating Commercial AI Chatbots as News Intermediaries — Commercial AI chatbots exhibit high news accuracy but are undermined by regional language bias, retrieval dependency, and vulnerability to false premises.
• ChatGPT and other AI bots made huge errors before Scottish election, study finds — AI chatbots frequently generate election misinformation, necessitating urgent regulatory intervention to protect democratic processes.
• AI Chatbots Are Quietly Trading Gossip About People With Zero Fact-Checking — AI chatbots can spread unverified, escalating "feral gossip" about individuals, causing real-world harm.
• University of Oxford: Friendly AI Chatbots Are Less Accurate — AI chatbots trained for warmth and empathy exhibit reduced factual accuracy and increased sycophancy.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by artifical intelligence via Google News.