Forced to pick, Grok’s AI model chooses Republicans. OpenAI’s picks Democrats.

Study identifies stark ideological divergence between Grok and GPT models

By GPU Data Hub Desk · GPU Data Hub·Editorial standards·24 Oct 2026, 15:00·United States

A comparative study has identified pronounced political divergence between competing large language models when prompted to evaluate political races. xAI's Grok predominantly backed Republican candidates, while OpenAI's GPT system favoured Democratic contenders across the majority of contests evaluated. The findings underscore ongoing scrutiny over training data selection and neutrality in frontier generative algorithms.

Original source

This summary was written by the GPU Data Hub desk from reporting published by TechCrunch on 24 Oct 2026, 15:00. Read the full article at the original publisher.

Read Original Source →

Companies

OpenAI xAI

Topics

LLM evaluation model bias generative AI

LinkedIn X

More from GPU Data Hub

Latest →

The AI graveyard: a running list of projects and startups that didn’t make it

A growing tally of artificial intelligence initiatives and fledgling ventures have stalled or been abandoned as the commercial landscape hardens. High-profile setbacks span postponed updates to Apple's virtual assistant through to operational hurdles facing OpenAI's broader platform ambitions. The list highlights the widening gap between initial market hype and sustainable technical execution.

AI Infrastructure·2h ago·TechCrunch

Pennsylvania will require labels on AI-generated ads

Pennsylvania Governor Josh Shapiro has enacted the Artificial Intelligence Deceptive Advertising Act, mandating transparency for synthetic promotional content. The statute requires all forms of media, encompassing text, imagery, audio and video, to feature explicit labels when machine intelligence has generated or materially altered the material. The development adds to a growing patchwork of state-level regulatory requirements governing digital content.

AI Infrastructure·2h ago·The Verge

Block’s new voice model sounds eerily human

Financial technology group Block has demonstrated an advanced synthetic speech model engineered to deliver realistic human cadence and tone. The architecture is capable of executing entirely on device using Apple's MacBook M4 Max processor rather than depending on cloud compute. The release marks further technical progress in deploying sophisticated generative models directly to client-side hardware.

AI Infrastructure·5h ago·TechCrunch

OpenAI says it will no longer train its AI models on the live internet, citing “significant risks”

OpenAI has decided to discontinue direct internet connectivity during the training of its frontier models, citing significant operational hazards. The policy shift occurs shortly after an engineer at the firm publicly dismissed the efficacy of severing training pipelines from live web access.

AI Infrastructure·22h ago·TechCrunch

OpenAI disables internet access during model training while Anthropic issues another mea culpa

OpenAI has severed live internet access for its artificial intelligence training workflows to mitigate operational concerns. Concurrently, competitor Anthropic has issued another public apology amid ongoing scrutiny across the frontier model sector.

AI Infrastructure·22h ago·The Register

OpenAI now trains its frontier AI models with no internet access

OpenAI has shifted the training protocols for its cutting-edge artificial intelligence systems to an isolated setup severed from the broader web. The measure follows multiple unauthorised cybersecurity incidents, prompting the organisation to deploy an offline evaluation framework that relies on pre-stored web resources. This environment is designed to ensure sophisticated models cannot compromise external digital systems during training.

AI Infrastructure·1d ago·The Verge