Gemini’s Deep Research now taps Gmail, Drive, and Chat when users allow it. Because the agent can fuse internal messages and files with web context, output quality rises along with privacy risk. This guide shows how to roll out safely: set consent norms, restrict high-risk teams, validate audit coverage, and keep DLP and labels active so Deep Research never reads more than policy permits.
ClickFix campaigns scale by coaching users to “fix” access issues with copy-paste commands. After the click, actors steal Microsoft 365 tokens or credentials and, in some cases, drop PureRAT for persistence. Break the flow by enforcing admin-only app consent, requiring phishing-resistant MFA, and blocking browser-to-shell chains. Investigate mailbox rules, token reuse, and OAuth grants whenever ClickFix pages appear in referral logs.
Smishing texts now target lost iPhone owners with fake “found” alerts and cloned iCloud pages. Verify only in Find My, keep Activation Lock tied to your Apple ID, and avoid links in messages. Set a SIM PIN, rotate your Apple ID password, and review trusted devices to prevent account takeover.
The UK is investigating whether Yutong electric buses can be remotely deactivated. Norway’s Faraday-cage tests and Denmark’s review raised alarms about SIM-enabled diagnostics and OTA updates. Operators should lock down telematics, broker OTA, and drill outage response now.
NVIDIA’s Jensen Huang says there are no active discussions to sell Blackwell chips to China. Because U.S. export controls bind shipments and Beijing restricts foreign accelerators in state-funded data centers, near-term access looks unlikely. This analysis explains what that means for procurement, security, and model roadmaps and how to design for heterogeneous accelerators without betting your budget on rumors.
Encrypted doesn’t mean invisible. Microsoft’s “Whisper Leak” shows a passive observer can classify AI chat topics by watching packet sizes and timing on streaming language models. Here’s how the side-channel works, who can exploit it, and which mitigations actually move the needle.
A China-linked crew still breaks in through legacy CVEs Log4j, Struts, Confluence, GoAhead then hides behind scheduled tasks and msbuild.exe to run memory-resident payloads. They probe domain controllers with DCSync, and they target misconfigured IIS by abusing ASP.NET machine keys to deploy TOLLBOOTH with SEO cloaking. Reduce risk by patching edge services, restricting LOLBAS on servers, rotating machine keys, and alerting on replication from non-DC hosts.
A cluster of malicious NuGet packages plants delayed logic bombs that crash apps, corrupt databases, and disrupt Siemens S7 PLCs. Remove the dependencies, rebuild from clean mirrors, and test with date-manipulation harnesses. Lock down registries, verify publishers, and stream build/runtime logs off-box to detect abuse early.
Texas sued Roblox, alleging deceptive child-safety claims, common nuisance, and predatory grooming patterns. The case spotlights how Robux incentives, off-platform chat, and moderation gaps expose kids—and what families can do right now.
Exposed Ollama APIs and a critical NVIDIA Container Toolkit flaw raise the stakes for AI infrastructure. Authenticate Ollama, close public 11434, and patch NCT to stop container escapes. Stream LLM and runtime logs off-box, rotate tokens, and validate least-privilege settings to keep model IP and GPU workers safe.