Claude, export controls and the case for keeping local models within reach On Tuesday, 9 June 2026, Anthropic released Claude Fable 5 and Claude Mythos 5. Three days later, neither model was available to anyone. The reason was not a…
At least 15 JetBrains plugins posed as AI assistants while stealing provider credentials. The campaign shows why developer tooling needs central governance and short-lived secrets. Security company Aikido reported a coordinated set of malicious plugins in the JetBrains Marketplace. The…
The WAF Blocked the Name. The Attacker Changed the Spelling. Update, 29 September 2026: Google’s 25 September report describes renewed ShinyHunters exploitation of CVE-2026-35273 across sectors. The group changed its request paths to bypass WAF rules protecting PeopleSoft’s Environment Management…
Unauthorised access to a World Food Programme registration system exposed data associated with roughly 600,000 Gaza households. In a conflict zone, ordinary identity and location fields can create extraordinary physical risk. The World Food Programme confirmed unauthorised access to its…
Cisco’s new infrastructure-security platform combines faster vulnerability research with runtime protections. It could reduce emergency patch pressure, provided temporary shields do not become permanent exceptions. Cisco announced an agentic platform for operating and defending critical IT infrastructure on 2 June.…
Google’s new AI Threat Defense portfolio aims to connect threat intelligence, exposure context and code repair. Its value will depend on whether organisations can govern the decisions between those stages. Google announced AI Threat Defense on 27 May, presenting a…
Microsoft’s multi-model agentic security system offers a credible glimpse of faster vulnerability discovery—but its strongest lesson is the combination of specialised agents, tools and human validation. Microsoft has presented a multi-model agentic security system that uses more than 100 specialised…
A coordinated compromise of npm and PyPI packages shows how trusted publishing, CI caches and maintainer identities can combine into an automated supply-chain attack. Security researchers reported a new “Mini Shai-Hulud” campaign on 11 May involving trojanised packages in the…
Two intrusions into Instructure’s Canvas platform show why a shared education service needs tenant-level visibility, tested fallbacks and careful treatment of attacker claims. Instructure detected unauthorised activity in Canvas on 29 April. On 7 May, the same actor obtained access…
Vercel’s April incident began with a compromised third-party AI application and ended inside internal systems. OAuth consent deserves the same scrutiny as a privileged integration. Vercel disclosed that an attacker gained unauthorised access to certain internal systems after compromising Context.ai,…
