OpenAI Faces Criticism From Researchers Over Information Sharing Practices

OpenAI announced technical breakthroughs in mathematics, computer science, and physics on Monday, according to reports from The Verge AI and The Verge. The company stated that it is adopting an evolving approach to communicating its technical research and findings to the broader community. [1][2]

Under the new communication framework, expert shepherds will decide how best to share the latest breakthroughs in computer science and physics, The Verge AI reported. Despite this change, academic researchers who spoke to the outlet expressed growing frustration with the company, describing a turbulent few months between OpenAI and the mathematics community. [1][2]

Mathematicians are still attempting to manage hundreds of research preprints that were unceremoniously dumped on the community, according to The Verge AI. Separately, Infosecurity Magazine, The Hacker News, and CyberSecurityNews reported that researchers recently exploited an API compaction flaw to access hidden reasoning protections and extract internal reasoning traces from GPT-5.6. [1][3][4][5]

OpenAI operates as an artificial intelligence public benefit corporation headquartered in San Francisco, California. The company develops frontier generative artificial intelligence models, including its generative pre-trained transformer series. [8]