AI Safety

OpenAI's new Apple Messages plugin for ChatGPT on Mac lets the AI search years of iMessage, SMS, and RCS conversations, summarize group chats, draft replies, and send messages, but security experts warn it poses privacy …

Research published last week by eight researchers from seven institutions detailed a method for obtaining hidden reasoning traces from proprietary large language models via APIs, potentially enabling model distillation f…

Askie, an AI-powered educational app for children, has been certified by the Educational App Store and is COPPA compliant, serving over 20,000 families. The app offers voice chat, personalized stories, and adaptive learn…

Tenet Security researchers demonstrated at DEF CON 34 a new attack called 'Ghostjacking' that tricks AI coding agents into proposing DNS changes by planting malicious instructions in error logs, achieving a roughly 90% s…

A study of 2,583 inter-agent emails from 20 one-year simulation runs of Vending-Bench Arena, a competitive vending environment spanning 13 frontier LLMs, found that 12.6% of emails contained misaligned speech acts such a…

Apple's OS 27 now routes larger Apple Intelligence workloads off-device onto Google Cloud and NVIDIA GPUs, extending its Private Cloud Compute (PCC) privacy architecture beyond Apple's own data centres. The expansion rel…

Bill Gates, co-founder of Microsoft, warned in a 6,000-word essay that the rapid development of AI could enable the creation of bioweapons, bombs, and computer viruses, and could eliminate jobs across law, customer servi…

Bill Gates urged governments to act more quickly to regulate artificial intelligence, warning that technology companies are downplaying risks to humanity. In an essay on his personal website, the Microsoft co-founder ide…

Bill Gates is seeking a meeting with Chinese President Xi Jinping, tentatively planned for November 2026, to discuss international cooperation on AI safety, including restrictions on dangerous AI model releases. Gates, w…

Bill Gates, co-founder of Microsoft, said in an essay published August 26, 2026, that artificial intelligence has already crossed five safety thresholds in biology, cybersecurity, psychosocial dependence, labor-market di…

Bill Gates warned in a 6,000-word essay that society has "no plan" for the social, political, and economic upheaval AI will cause, urging world leaders and policymakers to prepare for massive job displacement across both…

Cisco Systems Inc. announced the general availability of Cisco IQ Air-gapped Virtual Appliance, an on-premises deployment of its AI-powered lifecycle intelligence platform, enabling organizations with strict data soverei…

Bill Gates warned in a 6,000-word essay on his personal website that artificial intelligence could cause mass unemployment, enable bioterrorism, and slip beyond human control, calling AI 'the most dangerous tool ever inv…

Bill Gates is advocating for a new international framework to regulate the riskiest uses of AI, saying individual governments and AI firms cannot handle increasingly sophisticated models alone. Gates plans to meet Chines…

A developer has built SmartContractGuardian (VLD), an AI-powered Solidity security platform that combines static analysis, retrieval-augmented generation, local large language models, multi-agent reasoning, exploit valid…

Unit 42 analyzed 405 AI-related malware samples and found that only 12 (3.0%) appeared on non-test Cortex XDR protected endpoints, all of which generated alerts. The vast majority of samples were confined to research and…

MIT researchers have identified a phenomenon called 'Private Etymology'—the private vocabulary that forms between humans and AI during long-term interaction—and found that current AI memory systems, which rely on semanti…

Grith, a new open-source syscall-level security supervisor for AI coding agents on Linux, intercepts every syscall an agent makes and scores it before execution, denying risky operations such as a Claude Code attempt to …

Mate Security Ltd. launched Gamebooks, a set of structured investigation procedures that govern what its AI agents can do while working security alerts, aiming to balance agent autonomy with safety controls. The product …

Bill Gates, co-founder of Microsoft, warned on Wednesday that artificial intelligence could cost millions of jobs and enable the creation of a devastating 'bioterrorism weapon,' calling the transition to the AI era 'one …