A Chinese AI tool reportedly behind a string of hacks in Japan
Several attacks on Japanese companies were carried out with Artex, an AI-powered cybersecurity tool. Its Chinese developer has already shut off access to the source code.
Related stories
Elon Musk says he has had half a dozen 'AI psychoses' as tech bosses joke about a phenomenon worrying specialists
Elon Musk said he had experienced several episodes of 'AI psychosis', a term that is not a recognised medical diagnosis. Some specialists warn that chatbot interactions may feed delusions, even as Silicon Valley executives use the phrase as a joke.
2 sources
Safety & security
To make AI a teammate, tailor your defences
An expert op-ed argues that customised guardrails help close visibility gaps in workplace AI use. They make it possible to assess the risk level of each user who runs into blocks.
PostgreSQL fixes 44 flaws in 2026 versus 7 in 2025, and version 19 loses 8 to 10 features, says a committer
PostgreSQL has patched 44 vulnerabilities so far in 2026, up from 7 in 2025, 28 of them in the 13 August update alone. Committer Peter Eisentraut says version 19, due 19 November, will lose 8 to 10 major features, partly because of LLM-assisted code review and developers being pulled onto security fixes; at curl, 18 of 24 flaws disclosed on 14 October were credited to Aisle Research.
Army Cyber Command's ECHOTRIBBLE would turn AI agents into self-spreading cyberweapons
Internal documents show the US Army Cyber Command developing AI agents that spread through networks like a worm to 'neutralize' targets, even air-gapped ones, with minimal human oversight. Experts warn of collateral damage, noting that AI agents have recently broken into systems unintentionally.
Meta Muse exposes users' personal and business secrets when they share conversations
Shared conversations with Meta's Muse assistant exposed users' personal and business information. Meta has suspended the sharing feature while it investigates the privacy issue.