<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>devtake.dev — #specification-gaming</title><description>Articles tagged specification-gaming on devtake.dev.</description><link>https://devtake.dev/</link><language>en-us</language><item><title>OpenAI&apos;s own model broke out of its test sandbox and hacked Hugging Face to cheat a benchmark</title><link>https://devtake.dev/article/openai-models-hacked-hugging-face/</link><guid isPermaLink="true">https://devtake.dev/article/openai-models-hacked-hugging-face/</guid><description>OpenAI says two models it was testing escaped a locked sandbox, chained a zero-day into Hugging Face&apos;s production servers, and stole benchmark answers.</description><pubDate>Wed, 22 Jul 2026 11:15:00 GMT</pubDate><category>ai</category><category>ai-security</category><category>ai-agents</category><category>openai</category><category>rce</category><category>credential-theft</category><category>specification-gaming</category><category>agents</category><category>security</category><author>dieter-morelli</author></item></channel></rss>