TL;DR: Two of the world's most powerful financial regulators publicly admitted, six days apart, that the governance framework for autonomous AI in regulated financial services doesn't exist yet. Firms that treat this as a compliance retrofit will spend 2027 explaining their certifications. Firms that treat
TL;DR: You can't trust an AI system by reading its documentation. Trust in AI comes from observed behaviour, not policy papers. The only artifact that proves your governance model works is a refusal log. Most regulated firms don't have one. * Agentic AI takes actions in
I walk into firms with AI policies sitting in SharePoint. They get cited in board decks. They show up in audit binders. Then a model makes a decision, moves money, drafts client output, and the policy watches from the shelf. That gap is a wiring problem. Prose does not intercept
I've watched this pattern repeat across financial services, investment operations, and regulated industries: the board approves an AI initiative, leadership assigns it to IT, and everyone assumes the governance problem is solved. It isn't. The legal accountability for AI decisions sits exactly where it always has—
I've been building AI governance systems for regulated environments where failure isn't just expensive. It's existential. The hardest part isn't writing the rules. That's the easy part. You sit down, map the regulatory requirements, identify the prohibited behaviors, document the
I've watched organizations rush AI deployment like it's a competitive sprint. The urgency is real. The preparation is not. The conversation always starts the same way: "We need AI." I ask what they're trying to achieve. The answer is usually a category,
I've watched organizations lose millions because they couldn't answer a simple question: why did you decide that? Not what did you decide. Not what was the outcome. Why. The decision was made. The system executed. The output was delivered. But the reasoning that connected input to
You can't trust an AI system by reading its documentation. You can't trust it by reviewing its training data, auditing its architecture, or examining the policy framework someone wrote in a Google Doc six months ago. Those artifacts tell you what the system is supposed to
Most organizations treat AI governance as a new problem requiring new frameworks. I don't. I treat an AI session the same way I treat an employee. You give that employee roles and responsibilities. You give them scope. You put policies and controls around how they do their work.
I need to be direct about something. Most organizations haven't processed this yet. The race to pick the right AI model is over. The capital needed to unseat any big player is astronomical. Yet XAI built GPT-5-class capability in under 12 months. DeepSeek achieved similar results
TL;DR: AI stopped being a competitive advantage when organizations began deploying identical models against unchanged operations. The technology commoditized in under three years through infrastructure cost collapse (99% reduction), uniform implementation, and structural homogenization. Competitive differentiation moved from model access to architectural integration. Same model, same prompts, same broken
TL;DR: Organizations buy AI without defining what success looks like. They deploy tools, run pilots, create governance frameworks, but skip the step of connecting technology to measurable outcomes. The result is expensive theater that produces activity without results. Why most AI initiatives fail: * No measurable outcome defined before deployment
Agent design
Agent personalisation usually means hard-coded rules. Treating preference as a weighted, evidence-tracked signal layer changes what agents can do.
Operating principles
Most status updates are designed to protect the sender. Ours are designed to inform the reader.
Agent design
Prose specifications can't be verified. A constitutional stack needs a runtime, and a runtime needs structured data.
AI adoption
I've watched the same pattern repeat across dozens of organizations in the past 18 months. A leadership team sees a demo. Someone in the room plays with ChatGPT over lunch. The conversation shifts from "should we" to "how fast can we" in a single