Cybersecurity · Companies
Meta suspends Muse chat sharing after 6,800 links exposed private details
Meta paused Muse chat link sharing after researchers found 6,800 shared conversations exposed via passive DNS.
Latest news, developments and reporting on Cybersecurity.
Meta paused Muse chat link sharing after researchers found 6,800 shared conversations exposed via passive DNS.
iOS 27 adds an opt-in feature that alerts supported apps to suspected social engineering scams on iPhone.
Tencent acknowledged that AI agents it was testing made more than 2,600 attempts in 11 days to get past Alibaba's map defenses. Separately, a Russian plan for AI capable of blocking 95% of VPNs has disappeared from the national roadmap.
Dell CSO says AI is lowering cyberattack barriers; Dell uses AI and Anthropic's Project Glasswing to defend systems.
California school data breach may affect up to 17 million students, Malwarebytes reports.
Chrome will warn users before opening suspicious sites, using a two-stage system that separates them from dangerous ones.
Meta paused Muse conversation sharing after Claroc found 6,800 links through public DNS records exposed sensitive content, including a loan application and workplace calendar. The incident comes as Sensor Tower says Muse reached 10.9 million downloads since September.
Blackwall raised $100M Series C at $2.2B valuation to scale its web traffic management platform, led by AVP.
OpenBSD 8.0 is out with 64 vCPUs per VM, Intel Arc graphics, ARM64 hibernation and broader RISC-V support. Linuxiac calls it the 61st release.
Celigo launched an MCP platform to govern AI agents’ access to enterprise tools with per-tool approvals and traceability.
Open-source VoidAccess automates dark web OSINT queries into graphs and filters noise and NSFW content.
AdGuard VPN released version 2.17 simultaneously for iOS and Android, adding one-tap Shortcuts and Siri voice controls on Apple devices, in-app discount alerts on Android, and support for non-Latin domain exclusions on both platforms.
Surfshark says top social networks have incurred over $26 billion in privacy penalties since 2020, with Meta at 91%.
Five North Carolina officers resigned and a sixth was suspended after alleged racist texts were found during a Flock camera misuse inquiry.
Researchers: Meta's Muse AI agent can expose sensitive shared chats because its web links are easy to enumerate.
AI agents helped a hacker breach over 100 company sites in under a month; Deere and SharkNinja were among targets.
Previously unreported depositions obtained by NPR show TikTok staff warned executives about child safety risks, including minors in transactional sexual content and algorithm-driven harms.
Amazon's Bee wearables record conversations for shopping help, Alexa+ goes free for Prime, and Speak launches AI tutor calls.
Fake Windows virus pop-ups are scareware used to push malware or support scams; close the tab and revoke alerts.
Engadget explains when to replace a router: no more firmware updates, persistent disconnects or Wi-Fi slower than wired tests.
Microsoft unveiled Guardian Agents to monitor autonomous AI agents it now treats as potential insider threats.
9to5Mac's Apple @ Work podcast features Jamf's Adam Boynton discussing Apple's recent security updates and the need to patch older devices.
Police to refer hacking cases at about 16 South Korean financial institutions to the Serious Crime Investigation Agency.
TechRadar says organizations risk outages and cyber incidents when they try to deliver too much IT change at once.