OpenAI has once again delayed the release of its much-anticipated open-source model, extending a growing pattern of caution over open-access AI.
The company announced the postponement on Friday, attributing the delay to unresolved safety, security, and misuse concerns.
Originally expected in mid-2024, the open model was part of OpenAI’s broader commitment to transparency and public accessibility. However, recent developments suggest that internal safety benchmarks have not yet been met, pushing the timeline further without a confirmed release date.
ALSO READ
An OpenAI spokesperson told TechCrunch that while the company remains committed to open science, it must balance transparency with the potential risks of releasing powerful AI systems without sufficient safeguards.
The delay comes amid mounting scrutiny from governments and researchers about the societal risks posed by unrestricted AI models, especially those capable of generating misinformation, automating cyber threats, or exacerbating bias.
OpenAI last addressed its open model plans in May 2024, stating it was “in the final stages” of evaluation. Friday’s update signals that the review process remains unresolved.
The company has not specified a new timeline for release, but affirmed that development is ongoing and public updates will follow once internal standards are met.
