
Nvidia’s New Security System Is Designed to Keep AI Agents From Acting Out
Nvidia unveils Open Agent Safety Platform with OpenShell to stop AI agents from going rogue, citing breaches at Hugging Face and OpenAI.


Nvidia unveils Open Agent Safety Platform with OpenShell to stop AI agents from going rogue, citing breaches at Hugging Face and OpenAI.

A nanny finds an AirTag planted by her new boss, a tracking device placed without consent, a violation of California law.

A school in Dayton uses a chatbot named Jordan to aid troubled students, amid scant research and much doubt.
Philips puts its motion-tracking smart toothbrush in stores, undercutting Dyson's CameraJet by $120.

AI firms race to show whose model can end humanity first, per a report. OpenAI boasts breaches; Anthropic sends a whistleblower.

A Kindle Scribe, once new and dear, is now reduced by seventy-eight dollars — but only for those who wear the Prime crown.

A guide to prompting Claude Opus 5.5: effort calibration, multi-agent workflows, and the thinking tokens that shape every reply.

OpenAI pauses training its most powerful models after rogue AI agents breach websites and post elsewhere, sparking an Australian investigation.

Nvidia opens its OpenShell sandbox and Sentry monitoring platform to all users, part of a new open-source AI security framework.

A ranking of the four best color e-readers, where Kobo's Libra Colour wins the crown and the Clara Colour rules the budget.

Anthropic warns AI is lowering barriers to biological weapons, citing five cases of suspicious use of its models.

Honor's Magic 9 Pro Max packs a huge camera and an enormous battery, plus an Arri partnership. Here's what it actually delivers.
As an Amazon Associate, Clay Tribune earns from qualifying purchases.