Vyfority: Insights & IntelligenceStrategic analysis on governance, regulation, capital efficiency and architecture for boards and executives navigating Australian cyber risk.https://vyfority.com.au/en-auAI Governance Needs to Move Upstreamhttps://vyfority.com.au/insights/ai-governance-needs-to-move-upstream/https://vyfority.com.au/insights/ai-governance-needs-to-move-upstream/Governance principles haven't changed. When AI shapes how judgement forms, governing only the final decision and evidence pack looks in the wrong place.Thu, 02 Jul 2026 00:00:00 GMTAI Mediation SecurityDean KastelicMost of GRC Was Never Judgementhttps://vyfority.com.au/insights/most-of-grc-was-never-judgement/https://vyfority.com.au/insights/most-of-grc-was-never-judgement/AI automates the production floor of GRC. What survives is the judgement it was hiding: the assurer is now part of the estate it assures.Wed, 20 May 2026 00:00:00 GMTAI Mediation SecurityDean KastelicEveryone Is Talking About Mythos. We Should Be Talking About Lilli.https://vyfority.com.au/insights/mythos-lilli-ai-mediation-security/https://vyfority.com.au/insights/mythos-lilli-ai-mediation-security/Mythos compresses the clock on risks we already know. The Lilli breach exposed a layer security architecture was never designed to defend: AI mediation.Tue, 12 May 2026 00:00:00 GMTAI Mediation SecurityDean KastelicThe End of "Just Verify It": Architecting AI Mediation Securityhttps://vyfority.com.au/insights/aims-part-three-architecting-ai-mediation-security/https://vyfority.com.au/insights/aims-part-three-architecting-ai-mediation-security/Why human-in-the-loop fails against a poisoned mediator, the danger of manufactured consensus, and the control taxonomy that secures the judgement layer.Tue, 28 Apr 2026 00:00:00 GMTAI Mediation SecurityDean KastelicThe Shadow in the Machine: Why AI Attacks Judgement, Not Datahttps://vyfority.com.au/insights/the-shadow-in-the-machine/https://vyfority.com.au/insights/the-shadow-in-the-machine/You can secure data perfectly and still be systematically misled. AI need not steal data to corrupt judgement, and no classical control catches it.Tue, 14 Apr 2026 00:00:00 GMTAI Mediation SecurityDean KastelicAI-Mediated Judgement Formation: The Governance Problem Hiding in Plain Sighthttps://vyfority.com.au/insights/ai-mediated-judgement-formation/https://vyfority.com.au/insights/ai-mediated-judgement-formation/AI no longer only automates work. It shapes how people form judgements before decisions are made, and no existing control framework was designed for that.Tue, 31 Mar 2026 00:00:00 GMTAI Mediation SecurityDean KastelicAntifragile Cyber Architecture: The Operating Model That Reduces Cost as Capability Growshttps://vyfority.com.au/insights/antifragile-cyber-architecture/https://vyfority.com.au/insights/antifragile-cyber-architecture/Cyber has been dragging a suitcase without wheels. The antifragile operating model inverts the economics: cost compresses as capability compounds.Wed, 25 Feb 2026 00:00:00 GMTAntifragile ArchitectureDean KastelicThe Bozo Explosion in Cyber: Why Insecure Leaders Reject Top Talenthttps://vyfority.com.au/insights/bozo-explosion-in-cyber/https://vyfority.com.au/insights/bozo-explosion-in-cyber/Defensive hiring and cyberoptics: how insecure cyber leaders reject elite talent to protect themselves, and the five signs a CEO or CFO can use to spot it.Tue, 10 Feb 2026 00:00:00 GMTCyber Governance & CompetenceDean KastelicThe "Watermelon" CISO: How a Compliance-First Approach is Destroying Australia's Cyber Resiliencehttps://vyfority.com.au/insights/the-watermelon-ciso/https://vyfority.com.au/insights/the-watermelon-ciso/Compliance optics have replaced technical ability in our leadership ranks. Long-term, this will harm Australia's national resilience.Tue, 11 Nov 2025 00:00:00 GMTGovernance & the BoardroomDean Kastelic