Study Shows Grok AI Prefers Republicans, OpenAI GPT Prefers Democrats
A new TechCrunch AI analysis finds Grok's AI model heavily favors Republican candidates, whereas OpenAI's GPT shows a strong Democratic tilt.
Divergent Political Recommendations from Two Leading AIs
A study released by TechCrunch AI examined the political recommendations of two prominent language models: Grok's AI and OpenAI's GPT. The researchers evaluated the models across nearly 500 election contests, ranging from local to national races.
The findings indicate a stark contrast: Grok's AI overwhelmingly suggested Republican candidates, while GPT consistently leaned toward Democratic contenders. In more than 90% of the races, the two models offered opposite endorsements.
These results raise questions about the underlying training data and algorithmic biases that may influence political outputs. The study does not attribute the divergence to any intentional design but highlights the need for transparency in AI-driven political advice.
While the research stops short of prescribing solutions, it underscores the importance of scrutinizing AI behavior in the political arena, especially as such tools become more accessible to the public.
The full report can be accessed on TechCrunch AI's website, providing detailed methodology and race-by-race breakdowns.
Read also
OpenAI Pushes AI‑Driven Social Contract for 2028 Election
OpenAI’s Chris Lehane proposes an AI‑driven social contract to influence the 2028 U.S. election, aiming to benefit citizens while supporting OpenAI’s business goals.
Google Unveils Gemini 4 Carbon, Boosting Cybersecurity
Google's latest frontier model, Gemini 4 Carbon, promises notable improvements in coding, agentic tasks, and especially cybersecurity, ranking first on the AI Index v5.0 benchmark.
OpenAI Secures $11 Billion in Pending Funding Round
OpenAI is said to have raised $11 billion in a funding round that remains open, with closure anticipated shortly.
