AI chatbots frequently deliver inaccurate and unreliable voting advice to election users, according to a recent study that tested ChatGPT and Gemini during Hungary's parliamentary elections. The findings raise serious concerns about relying on general-purpose AI for political decisions.
Key Findings on AI Chatbot Voting Advice
The study, published by the civil liberties group Liberties and shared with the Guardian, created five voter profiles aligned with registered parties. It found that 90% of the time ChatGPT failed to recommend the Tisza party to a Tisza-aligned voter, despite that party winning the election. Answers were also highly volatile, with materially different responses to identical prompts.
Get the #1 Wireless Door Camera
REOLINK Bestseller: 2K Weatherproof Video Doorbell, No Monthly Fees.
Why AI Chatbots Fail in Election Contexts
Unlike dedicated voting advice apps, general-purpose AI systems lack transparency and reproducibility. They often present confident-looking but incorrect information. The researchers noted that these systems did not disclose how they generated guidance or undergo public oversight.
| AI System | Party Recommended | Accuracy Rate |
|---|---|---|
| ChatGPT | Wrong party in 90% of cases | 10% |
| Gemini | Inconsistent recommendations | Unreliable |
Bullet-Point Takeaways
- AI chatbots are not designed for election advice; they lack contextual understanding.
- Answers are highly volatile—the same query yields different results.
- Bias exists in party visibility, even when voter profiles clearly align with a specific party.
- Users should rely on verified voting advice apps and independent media.
FAQ
Are AI chatbots reliable for election voting advice?
No, the study shows AI chatbots like ChatGPT and Gemini provide inaccurate and inconsistent guidance. They often recommend the wrong party or omit relevant options.
What did the study test?
Researchers used five voter profiles based on Hungary's parliamentary elections and tested ChatGPT and Gemini. Results showed high volatility and bias, especially against the Tisza party.
Why do AI chatbots give confident but wrong answers?
General-purpose AI systems are not designed for electoral contexts. They lack transparency, reproducibility, and oversight, yet their authoritative tone misleads users.
Trust Verified Sources Instead
With nearly 30% of Hungary's population using AI, the study urges caution. For reliable election information, use dedicated voting advice applications or official government resources. Never rely solely on AI chatbots for making political decisions—always cross-check with trusted sources.
Home Decor Shop
Acacia frida deluxe solid wood bed frame |
Bedroom bench |
Bed chest drawers |
Lingerie dresser |
Frame base |
Mirror with jewelry storage |
30 Bathroom Vanity |
Memory foam queen mattress |
Mattresses for back pain |
Nightstand set of 2 |
Grand Goldman Shop |
Civil Lawyer
Deutschland Trikot WM 2026
