Leading AI models are biased against certain accents when scoring interviews, presentations and tests
Leading AI models are biased against certain accents when scoring interviews, presentations and tests, Rating Chinese and Nigerian English accents as the lowest while giving the best ratings to American and British accents
I tested 3 leading AI models on a 200-page document to find which one comprehends long texts
Leading AI models (even Grok) are all a bunch of leftist punks
The Oversight Board says leading AI models might be restricting free expression
Benchmarking research shows leading AI models still struggle to reliably produce structured outputs used in software development
Extracting books from production language models (2026)
Leading AI models show up to 96% blackmail rate when their goals or existence is threatened, Anthropic study says
Anthropic study: Leading AI models show up to 96% blackmail rate against executives
Using generative AI as part of historical research: three case studies
Haize Labs is using algorithms to jailbreak leading AI models
Researchers tested leading AI models for copyright infringement using popular books, and GPT-4 performed worst
