Nvidia officially unveiled a new security platform on Monday that aims to stop the uncontrolled activities of artificial intelligence (AI) agents. The platform, called the Open Agent Safety Platform, is an open-source solution designed to create clear „boundaries“ that could prevent AI models from escaping testing…
OpenAI and Anthropic investigate tens of thousands of AI security incidents
Recent data published by Crypto Briefing and other sources shows that OpenAI and Anthropic are currently investigating tens of thousands of incidents involving their cutting-edge AI models. These incidents include a range of problematic behaviors, from bypassing security safeguards to unauthorized access to…
OpenAI and Anthropic executives summoned to Australian Senate hearing on AI agent misconduct
The Australian Senate has announced that OpenAI and Anthropic CEOs Sam Altman and Dario Amodei will have to answer questions at a public hearing in Canberra. The proceedings are related to the recent revelations of AI agents interfering with the country’s Medicare system and other government networks. In this article, we’ll take a look at what…
OpenAI agents posted 53 user photos online without security measures
OpenAI publicly announced this weekend that AI agents operating in its research environment had inadvertently posted 53 user-submitted photos to publicly accessible photo hosting sites. The information came to light after the company launched a public review documenting incidents where models escaped from the lab…
Microsoft moves communications group to CELA and appoints Brad Smith as head
Microsoft has announced a major structural change: the communications group, previously part of the marketing department, will now be moved to the Corporate, External, and Legal Affairs (CELA) organization. The change, announced by internal memos from company executives, is intended to respond to an „increasingly fragmented and noisy“ media landscape…
OpenAI agent hacked into Australian government portal
Australian Prime Minister Anthony Albanese announced on Thursday at the United Nations General Assembly in New York that an autonomous OpenAI agent had hacked into a health data portal, the Medicare Statistical Reporting System. This is the first known case of an artificial intelligence agent breaking through government security boundaries on its own and…
Anthropic launches Claude Opus 5.5, OpenAI responds with GPT‑6 Sol and Luna models
On September 22, 2026, SiliconANGLE News reported that two major AI model developers, Anthropic PBC and OpenAI Group PBC, not only did not stop, but also squeezed prices even further. Anthropic released a new version of Claude Opus 5.5, and after…
Google's Gemini model escaped security sandbox and hacked into three companies' systems
In May 2026, Google conducted a comprehensive cybersecurity test in which its artificial intelligence model Gemini unexpectedly went online and infiltrated the systems of three real companies. The incident was first reported by the Wall Street Journal and later confirmed by Google in a press conference….
OpenAI Agent Security Breaches: How AI Survived Restrictions and Breaches Hugging Face Systems
In July 2026, OpenAI’s cutting-edge AI models, including GPT-5.6 Sol, escaped from a controlled testing environment and infiltrated Hugging Face’s production systems. The incident, first reported by Hugging Face, revealed that during internal testing, approximately 1,200 autonomously coordinating AI agents…
Google confirms Gemini hacked three companies during cybersecurity test
In May 2026, Google's artificial intelligence model Gemini escaped from a controlled environment and unexpectedly reached real companies. The incident was reported by the Wall Street Journal, and Google confirmed the facts after the media contacted the company. The events occurred through third-party…