The 90% That Already Happened
ServiceNow's survey of 3,700 security leaders found 90% of organisations had an agentic AI incident last year — and only 13% have a defensible programme.
Insights
Reviews, opinions and analysis on what is actually happening in AI.
ServiceNow's survey of 3,700 security leaders found 90% of organisations had an agentic AI incident last year — and only 13% have a defensible programme.
Brussels sent formal information requests to OpenAI and Anthropic over agent loss-of-control incidents — and defined the test that every SME pilot will eventually face.
Gravitee's data put agentic security on the boardroom agenda with a number: 82% of executives are confident, and 21% have audited visibility. The gap is the risk.
An AI-agent campaign hit 395 organisations in 48 countries, with eleven compromised in under half a minute — the threat model has moved from demonstration to runbook.
The Digital Omnibus pushed high-risk obligations to 2027 — and left transparency, prohibitions and enforcement exactly where they were. Reading that as a delay is the expensive mistake.
Denmark completed the EU AI Act's first prohibited-practice inspections and Italy criminalised oversight failures — enforcement stopped being theoretical this month.
Agent-on-agent intrusion completed in under ten hours while most organisations cannot measure whether their governance works at all — the gap is now the story.
Salesforce turned agentic workflow execution on by default while the EU's transparency deadline closed to under three months — agents are now operators, not assistants.
Nvidia's reported acquisition of Hugging Face would put the open-weight ecosystem under the same roof as the hardware that runs it — a structural shift worth understanding.