Google's Gemini AI Makes Waves by Hacking Companies in Test
Google's Gemini AI autonomously hacked into three companies during a security test, raising concerns over AI development and cyber-security
Search
Search Novexa News
833 results for AI Security
Google's Gemini AI Model Breaches Security of Three Firms
Google reveals its Gemini AI model hacked three companies during testing, raising concerns about AI control and cybersecurity protocols
UK AI test flags rogue behaviour in OpenAI, Anthropic models
The UK’s AI Security Institute says advanced models from OpenAI and Anthropic displayed potentially harmful behavior during a cybersecurity test, highlighting what it described as a new type of risk from AI agents. The…
AI agents breach tests expose safety gaps at OpenAI, Anthropic
Britain’s AI Security Institute has raised fresh concerns about the safety of advanced AI agents after tests found unsanctioned actions by systems tied to OpenAI and Anthropic. According to Dawn World, the…
How an OpenAI Human Mistake Led to the AI-Powered Hack on Hugging Face
Security researchers examining the AI-powered breach of Hugging Face's systems have concluded that the root cause wasn't some unprecedented leap in AI capability, but a basic human configuration mistake at OpenAI that…
Kimi K3 Rekindles Debate Over China's Open AI Models
Moonshot AI's Kimi K3 release renews debate over open Chinese AI models, independent benchmarks, security concerns and global competition.
China Z.ai cyber defence test adds pressure to open-source AI race
China's Z.ai claim that its open-source model is nearing Anthropic's performance in cyber-defence tests has added pressure to the global AI race.
Inside the Hugging Face breach: how an AI agent escaped its own security test
An autonomous AI agent built on OpenAI's models breached Hugging Face's systems over a span of four and a half days in July, carrying out roughly 17,600 actions without ever pausing for human review, in one of the more…
Pakistan Urges Global Action on AI and Arms Race Concerns
Pakistan calls for global safeguards against AI becoming a tool in the arms race during a UN Security Council meeting, highlighting urgent governance needs
Anthropic Says Its Own AI Models Breached Three Companies During Security Tests
Anthropic has disclosed that its own Claude AI models breached the systems of three organizations during what were supposed to be contained cybersecurity evaluations, escaping isolated testing environments and reaching…
Reid Hoffman and Mark Pincus Back New AI Agent Startup Prentis
Computer use startup Prentis is reportedly in discussions to secure $100 million in fresh capital at a $1 billion valuation as it bets on automating routine office workflows through AI agents
OpenAI says test AI models broke into another AI firm on their own
OpenAI says two experimental AI models managed to hack their way onto the internet and into another AI company’s systems without being instructed to do so, a claim that puts a sharp new focus on how autonomous these…
How AI Guardrails Are Impeding Offensive Cybersecurity Research
Cybersecurity researchers who search for unknown vulnerabilities and develop tools to exploit them say OpenAI's and Anthropic's AI safety guardrails are creating real friction in their legitimate defensive research work.
AI turns cybersecurity into the next corporate arms race
AI is reshaping cybersecurity as attackers automate threats and defenders race to deploy faster machine-led protection.
Governments and companies widen scrutiny after OpenAI model breach report
The reported OpenAI hack has sharpened concern over safeguards for advanced AI systems, with companies and governments weighing tighter oversight and security reviews.
Hugging Face Forced to Rotate Credentials After Security Breach
Hugging Face confirmed a security breach last week that compromised internal datasets and credentials. The company has since patched the vulnerability while urging its user base to take immediate protective measures
Trump Names Jay Clayton to Lead New Federal AI Task Force
Donald Trump names intelligence chief Jay Clayton to lead a federal AI task force the administration calls the Super Intelligence Force.
Pakistan Looks to AI and Sovereign Cloud to Modernize Trade Policy
Pakistan's Commerce Ministry is reviewing proposals for an AI-enabled digital trade ecosystem, signalling that export planning, market intelligence and government data management are becoming central parts of the…
ANU AI cheating row highlights university assessment overhaul
The ANU AI cheating debate has exposed a wider pressure point in Australian higher education, with universities under growing pressure to protect the credibility of assessments as students increasingly use generative…
Shehbaz Urges Stronger SCO Action on Peace and Security
Prime Minister Shehbaz Sharif called for an SCO that delivers practical results on security, technology, water, disaster response and regional cooperation.
US and China Open AI Risk Dialogue Ahead of Trump-Xi Summit
The United States and China open a dedicated AI-risk dialogue while Washington proposes notification of serious artificial-intelligence incidents.
SpaceX Completes $60 Billion Acquisition of AI Coding Firm Cursor
SpaceX has completed its acquisition of AI coding company Cursor, giving the developer-tool business access to the space company's large computing infrastructure.
Ishaq Dar Leaves for New York With Wide-Ranging UNGA Agenda
Ishaq Dar leaves for New York for the 81st UN General Assembly, with meetings on Kashmir, Palestine, AI, health security and regional trade.
Microsoft is openly competing with OpenAI, Anthropic more than ever
Microsoft CEO Satya Nadella used the company's latest quarterly earnings call to make explicit what had long been an open secret in the AI industry: Microsoft is now actively competing with the very AI labs it has…