Reinforce Identity and AI Controls to Match Machine-Speed Threats

What happened

A survey shows frontier AI is enabling attacks that move at machine speed and most Australian and New Zealand organisations cannot detect or respond at that tempo. The study highlights large blind spots in monitoring and that only a small fraction can detect compromise within minutes outside business hours. Watch whether tool adoption and playbook automation close the gap or whether detection/response capability continues to lag

Why the category manager should care

Treat machine-speed attack risk as an operational requirement: procurement must demand measurable detection and automated response capabilities from suppliers

Key facts

• Survey of more than 850 IT and cybersecurity professionals across Australia and New Zealand
• Low share of organisations can respond at machine speed; significant unmonitored areas reported

What happened

Visa released updates to an open-source vulnerability remediation framework that extends discovery into closed-loop remediation and validation. The vendor says some remediation timelines can shrink from weeks to hours and supports multiple AI models, making integration and model choice a procurement issue. Watch proof-of-concept results: closed-loop tooling can save time but may expose integration, auditability and model-hosting costs

Why the category manager should care

Treat remediation frameworks as integration projects that carry ongoing hosting and support obligations, not purely a one-off purchase

Key facts

• Visa extended its Vulnerability Agentic Harness from discovery into remediation and validation
• Framework is open-source and supports multiple AI models with optional real-time progress views

What happened

Mimecast is positioning to handle agentic AI risk, noting agent multiplicity and large numbers of AI-native applications and services inside customer estates. The vendor claims high automated remediation rates but also flags that most risk remains user-tied and that human oversight remains necessary. Watch vendor validation on shadow AI discovery and concrete failure modes for agentic actions in customer environments

Why the category manager should care

Validate vendor remediation claims in your environment and require defined human-in-loop thresholds for critical actions

Key facts

• Vendor reports large counts of AI-native applications and AI-enabled services in customer est
• Vendor claims majority of detected threats are auto-resolved with human oversight for a small

What happened

A survey of mid-to-large organisations shows a gap between executive confidence and the reality of authentication modernisation, with many systems not recently updated and identity work deprioritised. The mismatch creates ongoing exposure from stale accounts, over-permissioned users and duplicated administration across physical and logical access. Watch procurements that separate logical and physical access — combined sourcing usually reduces duplicate admin and improves oversight

Why the category manager should care

Prioritise identity modernisation in sourcing to reduce repeat incidents and supplier fragmentation between physical and logical access

Key facts

• Survey of 500 IT and security leaders highlighting modernisation gaps
• More than half have not significantly updated authentication systems in several years