André Afonso Silva·5d agoReduced, Not SafeIn October, the warning about a shared base could see down to configuration settings. It vouched for one step. The precaution needed…
martino.agostini·Oct 13AI Risk Governance: The Ultimate Leverage Point Is Decision AuthorityWhy interconnected AI risks require organizations to move beyond risk identification toward systemic governance, enforceable controls, and…
André Afonso Silva·Sep 22Agreed Upon ScopeWhat an independent examiner’s own investigation template says about who sets the borders of an examination
André Afonso Silva·Sep 17Five Factors, One EvaluatorWhat the AI Security Institute’s incident report shows about who controlled the conditions
André Afonso Silva·Sep 10Nobody Measures IndependenceThree layers of the AI supply chain rest on the same assumption. In July, one of them was found to have stopped holding, by accident.
The Insider Educator·Sep 7I Reverse-Engineered the Global Collateral Circuit.The hidden chain connecting sovereign bonds, repo, FX swaps, dealer balance sheets, central-bank facilities and forced liquidation
André Afonso Silva·Sep 3The Door Before DeploymentEvery frontier model passes through an evaluation before it reaches you. This summer, four organisations reported that those environments…
Paul Afraides·Aug 23Wall street's opacity problem is a political choice, not an accidentThe 2010 Dodd-Frank Act gave regulators the tools to force transparency, but the SEC has used its discretion to let opaque securities…
The Modern Lens·Aug 23Three stories. One lesson: Why Uniformity Always FailsThree case studies on how over-optimization creates systemic fragilityA response icon1A response icon1
