Crooked Bough News Round-up

— Okta debuts agentic lifecycle management tools, AI kill switch

 

at 9/22/26 3:00pm

In the wake of a string of incidents involving badly-managed AI agents, security firm Okta enhanced its year-old agentic services proposition with new capabilities to securely manage agentic lifecycles.

— Zero-trust was built for people. Nobody told it about agents

 

at 9/22/26 7:48am

The concept of zero-trust needs no introduction, but in the new world of AI agents, it’s a model that now needs to evolve.

— Data dive: Mapping UK police forces’ hyperscale dependence

 

at 9/22/26 6:20am

Computer Weekly used public DNS records to map which outside companies Britain’s 48 police forces connect to – and found a service that has quietly standardised on one US hyperscaler

— Splunk helps US research uni embrace ‘student-powered’ security

 

at 9/21/26 10:30pm

When she was asked to let IT students run its SOC, New Jersey Institute of Technology CISO Sharon Kelley ran a mile. Find out why she changed her mind, and how Splunk is helping her raise a new generation of cyber analysts

— EU states failing to exchange information on cross-border cyber security attacks

 

at 9/21/26 6:30am

European Union urged to deploy real-time European cyber alert platform to share information with member states on cross-border cyber attacks

— US cyber agency endorses ‘decoy’ tactics

 

at 9/18/26 6:28am

Official US guidance suggests organisations consider using cyber decoys to strengthen their detection and response capabilities.

— Pacing the AI Frontier: the risk of artificial unintelligence

 

at 9/18/26 12:38am

The most immediate risks stem not from artificial superintelligence but from artificial unintelligence

— Home Office challenged on ‘farcical’ secrecy over Apple ‘backdoor’ order

 

at 9/17/26 7:42am

UK government argues that national security would be damaged if it departs from ‘neither confirm nor deny policy’ on Apple ‘backdoor’ notice

— Advice for CIOs on AI’s 'existential threat'

 

at 9/15/26 11:05pm

Statements by OpenAI chief Sam Altman, following an essay by Anthropic chief Dario Amodei, raise concerns over the safety of frontier models

— CW@60: What Doom taught me about networking, security and trust

 

at 9/15/26 10:00pm

How John Kindervag went from setting up LAN parties for playing Doom at the office, to creating the zero-trust model of cyber security

— Cisco users urged to patch email gateway flaw

 

at 9/15/26 10:09am

Cisco warns defenders to patch a critical vulnerability in its Secure Email Gateway appliance that may be used by attackers to gain root privileges.

— Splunk Conf 26: Splunk hopes to build confidence in agentic AI

 

at 9/15/26 7:47am

With trust in AI agents close to rock bottom in the popular imagination, Splunk and Cisco unveiled a series of platform enhancements designed to enhance observability and confidence in agentic AI

— UK, US and Netherlands warn of Iranian state spyware campaign

 

at 9/15/26 7:18am

Cyber attackers linked to Iran’s Ministry of Intelligence and Security are targeting opponents and opposition groups with Windows spyware  

— CW@60: Don’t forget your toothbrush - guarding sources, exposing the secret state

 

at 9/14/26 10:00pm

Computer Weekly's Bill Goodwin has seen both sides of the courtroom during a career spent exposing the surveillance state and protecting those who wish to let the world know its secrets

— UK rolls out digital ID for alcohol sales

 

at 9/14/26 1:01pm

Changes to the law that will enable the retail and hospitality industry to introduce digital identity checks for alcohol sales are being rolled out

— UK lawmakers call for new law to protect human rights from AI systems

 

at 9/14/26 8:26am

MPs and peers call for the UK to create a statutory AI oversight body with powers to test and evaluate high-risk uses of AI and to prevent the deployment of AI that poses risks to human rights

— UK cyber shortfalls: A skills gap and/or a gender gap?

 

at 9/14/26 12:49am

There is both a skills and gender gap occurring in UK cyber, says Code First Girls – something it is looking to remedy with an innovative partnership with Lloyds Banking Group

— GDPR wasn't designed for AI and that's a security problem

 

at 9/11/26 4:06am

The Security Think Tank explores the intersection of GDPR and artificial intelligence, considering how data protection standards in the UK and Europe are changing in this new paradigm.

— Anthropic calls for ‘verifiable effort’ to control frontier AI

 

at 9/9/26 11:15pm

After Claude Mythos circumvented guardrails in July, Anthropic now wants an industry effort to control the pace of frontier model development

— Why AI agents cannot be trusted to secure agentic AI yet

 

at 9/9/26 3:10am

The idea of using AI agents to secure AI agents seems compelling, but in doing so, we risk unnecessarily expanding our attack surface.

As of 9/23/26 2:39pm. Last new 9/22/26 7:48pm.