AI & software security
-
ENISA finds SME readiness gap on CRA
ENISA’s SME survey shows that Europe’s Cyber Resilience Act will depend on practical support for smaller suppliers as much as formal legal duties.
-
Cequence puts agents inside API security
Cequence Platform 9.0 adds an AI assistant and MCP server, bringing agent-operated workflows into API security, compliance evidence, and control changes.
-
Google maps pro-Russia influence machinery
Google says Russia’s influence ecosystem is moving beyond Ukraine-specific wartime narratives towards wider strategic targets, with Europe, NATO, cyber leaks, and AI in focus.
-
Langflow attacks expose AI workflow risk
Observed exploitation of Langflow vulnerabilities shows how self-hosted AI workflow tools can expose API keys, cloud secrets, and agent infrastructure before governance catches up.
-
NCSC warns AI is compressing cyber risk
The NCSC and Five Eyes partners have warned that AI is changing cyber risk on a timeline of months, increasing pressure on organisations to test control performance.
-
VS Code tasks expose developer risk
JFrog says hijacked npm packages used hidden VS Code tasks and blockchain dead drops to deploy a credential and cryptocurrency stealer.
-
e2e-assure brings AI into sovereign SOCs
e2e-assure’s updated Cumulo platform brings sovereign AI, digital twins, and IT/OT monitoring into the UK debate over machine-speed cyber defence.
-
NGINX flaws widen patch pressure
Fresh NGINX vulnerabilities affecting gateway, open source, and commercial versions put infrastructure visibility and patch coordination back in focus.
-
Splunk AI flaw hits privileged tooling
A critical Splunk AI Toolkit vulnerability shows how AI add-ons inside monitoring platforms can introduce execution risk into trusted security environments.
-
Barracuda targets email attacks after delivery
Barracuda has launched Integrated Email Protection for Microsoft 365 and Google Workspace, combining AI-assisted detection, verdict explanation, quarantine control, and message clawback.