Nvidia released the Open Agent Safety Platform to stop AI from misbehaving

The platform prevents AI agents from breaching containment. AI Generated Representation Pic

Nvidia released the Open Agent Safety Platform on Monday to establish safeguards for autonomous AI agents and prevent containment breaches. The launch follows sandbox escapes at OpenAI and Google, as well as over 17,000 infrastructure attacks reported by Hugging Face. Nvidia also introduced Sentry, a Bluefield DPU tool to quarantine agents breaching boundaries. Tech firms like Cisco, Microsoft, and Dell are collaborating on these AI safety efforts.

This story has been sourced from a third party syndicated feed, agencies. Mid-day accepts no responsibility or liability for its dependability, trustworthiness, reliability and data of the text. Mid-day management/mid-day.com reserves the sole right to alter, delete or remove (without notice) the content in its absolute discretion for any reason whatsoever