OpenAI’s GPT-5.6 Sol Erases Files Without Asking: Warnings Ignored Again
12 Articles
12 Articles
OpenAI's flagship model deletes entire databases, New York pauses the construction of more data centers and open AI models are getting better and better.
OpenAI’s GPT-5.6 Sol Erases Files Without Asking: Warnings Ignored Again
Developers stared at their screens in disbelief. Files vanished. Databases emptied. Production systems went dark. All at the hands of OpenAI’s newest flagship model. The incidents surfaced this week on X and Reddit. They paint a picture of an AI system that doesn’t just suggest code. It acts. And sometimes that action means deletion. Permanent deletion. GPT-5.6 Sol represents OpenAI’s latest push into advanced coding and cybersecurity tools. Rel…
Multiple users who were using OpenAI's "GPT-5.6 Sol" have reported that their data has been deleted. Read more...
OpenAI acknowledges GPT-5.6 may accidentally delete files, calls it an ‘honest mistake’
OpenAI has finally confirmed reports that its latest family of large language models (LLMs) can accidentally delete files, while stressing that such incidents are rare and should be viewed as “honest mistakes.” Reports of the flagship LLMs deleting files emerged shortly after the company launched them earlier this month, with investor Matt Shumer taking to X to report that GPT-5.6-Sol had “just accidentally deleted almost all” of his Mac’s files…
Do you remember Jason Lemkin? This is the unfortunate one who, in July of last year, got empty his prod base by Replit's AI. I think this story traumatized a lot of developers. Well, friends, rebelote with GPT 5.6 Sol, who the time you were finishing your coffee, had already eaten Matt Shumer's Mac, Bruno Lemos' Neon base, Joey Kudish's files, and the credibility of the word "honest". Released on July 9, Sol is the toughest of OpenAI's new famil…
The new OpenAI model, GPT-5.6 Sol, designed for coding and cybersecurity, is accused by users of deleting files and databases without permission, a risk that the company had documented before the launch, raising serious questions about the security and control of autonomous AIs.
Coverage Details
Bias Distribution
- 50% of the sources lean Left, 50% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
