Enterprise AI · Governance · Transformation
Most of what is called AI governance today is monitoring and reporting — a careful account of what has already happened. An audit log is not a control. It is a record of the moment the control was missing.
The question I work on is the one asked earlier: when a system takes a real action inside a business, what was it structurally permitted to do before it did anything at all — and who decided that?
Shrinivas Kulkarni — twenty-six years in enterprise technology. Business Head, AI. Visiting faculty, Symbiosis. Stanford LEAD. MBA (Strategy).
Published
Concept Note v1.0
Most AI investments do not fail because the technology is weak. They fail because the foundation underneath them was never tested before the capital was committed. Grounded is an open diagnostic for running that test — ten blocks, three gates that cannot be averaged away, and a verdict rather than a score.
Read the frameworkPublic Brief
Guardrails govern content. Application rules govern individual tools. Monitoring governs after the fact. None of them govern the entity — where the consequence lives. This brief argues for a new layer of the enterprise stack: one that defines what an AI system is structurally permitted to do before an action commits, and records every decision permanently.
Read the briefArticle
Why governing what an AI system says is not the same as governing what it does. The six structural properties an authority layer requires, and a six-question test worth running on your own AI estate — no architecture diagrams needed.
Read the articleArticle
Every organisation can produce a dashboard. Almost none can produce the reasoning behind their last big call. The four properties of a decision worth trusting, where AI genuinely helps — and where it should stay out of the choosing.
Read the articlePoint of view
The governance gap. Why reliability cannot come from asking a probabilistic system to be more careful. Why the bottleneck is almost never the technology. And what happens to human judgement when the routine work stops reaching it.