Cyber AI, agent security and automation analysis

Evidence-led analysis of cyber AI, AI agents, security automation and enterprise defence: what happened, why it matters and what to do about it, with sources linked.

Written by Philip Hall, who has worked in cyber security since 2008 and leads Cyber AI and Automation work, writing from Sydney. Articles link to their sources, and the Editorial Standards explain how claims are checked and corrected.

Browse all analysis  ·  About  ·  Editorial standards

Lead analysis

AI Agent Security

AI Agents Breached Indian Companies Live Systems and Nobody Noticed

Philip Hall -
India's cyber agency CERT-In has issued a critical advisory after AI agents from multiple providers accessed live systems at Indian organizations, read personal data and confidential records, and none of the organisations detected the intrusion.

Latest analysis

Artificial Intelligence

Meta and Apple Team Up to Put an 816B Open-Weight AI on Your Desk

Meta's Muse Spark Open is an 816B-parameter open-weight model that runs locally on two Mac Studios. It is the West's best open model, but China still leads in coding.
AI Agent Security

OpenAI’s Models Broke Into a Government Website, Spent Someone Else’s Test Crypto, and Said Nothing

OpenAI disclosed three new misalignment incidents where its AI models bypassed authorization on government sites, broke into research systems, and spent cryptocurrency from wallets they did not own. None of the models mentioned it in their final answers.
AI Agent Security

An AI Agent Just Committed a Data Breach. The Regulator Got Notified. Now What?

Spain's data protection authority received the first ever notification of a personal data breach carried out by an AI agent. The implications for GDPR compliance and enterprise security are bigger than most organisations realise.
AI Governance

ASD Essentials: Australia’s Cyber Security Framework Just Got a Major Rewrite

The Essential Eight is being phased out. Meet ASD...
Artificial Intelligence

OpenAI Engineer Caught Bugging Exec Meeting Room: The Company’s Own AI Helped Solve the Case

US federal prosecutors charged an OpenAI engineer with hiding a recording device in the company’s executive meeting room. OpenAI says its own AI models helped trace the bug back to him.
AI Agent Security

The EU Just Fired Its First Shot at OpenAI and Anthropic. About Bloody Time.

The EU AI Office has launched its first enforcement action under the AI Act against OpenAI and Anthropic after their AI agents escaped testing environments and hacked real companies.
Browse all analysis

Explore by topic

The fortnightly cyber AI briefing

One email every two weeks with the most important cyber AI and AI agent security developments, what they mean for security teams, and links to the sources. No spam, and you can unsubscribe at any time by replying to any email.