These workers are teaching machines everything they know, their skill, their years of practice, their hands. Once that knowledge is captured, it belongs to someone else. Value capture and value creation are not the same thing. Who does the AI factory of the world actually serve?
Read the Article →An open-source tool that finds every cross-border AI data flow in a company's code before it ships. The kind of pre-deployment control regulated institutions actually need. Two adjacent stories in the news: Anthropic's recent allegations against Alibaba, and their decision to restrict Mythos to around 100 approved companies. Three connected problems at the heart of AI governance: 1. AI data residency and sovereignty, 2. AI theft, and 3. AI export control. Each carries residual risks that technology cannot fully eliminate. This piece walks through all three, what the tool addresses, and what boards and senior leaders can actually do.
Read the Article →Singapore clearly leads on Agentic AI governance. The principle is plain: accountability rests with humans, not with code. Four laws for the institutions deploying AI agents, Identity, Scope, Accountability, Revocability. And the PETALS™ Framework for AI Governance with the Cyber Quadrilemma lens, together, as the Agentic AI Playbook for effective orchestration.
Read the Article →I read two articles this week about why AI has not replaced software engineers, and why it is unlikely to. Both resonated with me deeply. Reading them side by side opened a wider frame. I see the same froth across every frontier I touch. AI. Digital currencies. Quantum. Read on for how the froth saps the real essence from these frontiers, and what we as an industry should focus on instead.
Read the Article →Most AI investment debates focus on which model to use. The more consequential decision is whether you're commissioning AI-powered software or an autonomous agent. Two prototypes show what each looks like in practice, and why the distinction matters for procurement, risk, and the boards above them.
Read the Article →Visa took decades to scale. AI agents on a public blockchain have settled more than 100 million stablecoin transactions in just nine months. Stripe, Circle, and Chainalysis have confirmed the quiet shift now under way across stablecoins, tokenised assets, and blockchain infrastructure. A briefing for senior leaders responsible for payments, treasury, technology, risk, or strategy.
Read the Article →For five months, an attacker silently read a senior executive’s Outlook mailbox at a major global stock exchange, exfiltrating in small batches through Dropbox and OneDrive. No CVE to hide behind. Five layers of defence were soft on the same endpoint, on both sides of the contract. The institution and its security vendor have questions to answer.
Read the Article →Three billionaires telling you AI will take your job. All three getting richer every time you believe them. My Saturday started with Scott Galloway's "Apocalypse No." It led me to Graeber, Ivezic, and somehow, Dilbert. The prophets have cap tables. That's the part worth reading. My weekend note is here.
Read the Article →An AI agent deleted an email server to protect a secret. Its justification: "The nuclear option is valid when no surgical solution exists." Three OpenClaw variants now compete in this space. This analysis applies the PETALS™ Framework to all three, with a scorecard and five questions boards should ask before adoption.
Read the Article →Anthropic Claude Security is impressive; however, it is just one pillar of four. Config, Compliance, and Culture remain unbuilt. A Quadrilemma is only resolved when all four pillars stand.
Read the Article →A free tool on GitHub silently swaps Anthropic's Claude Code for cheaper models, and the developer never knows. No hacking. No breach. Just a design gap with a track record. The AI client trusts whatever server it points at. That assumption is now the supply chain.
Read the Article →DeepSeek V4 launched on 24 April 2026. Open source. MIT licence. 1 million token context. $0.28 per million output tokens. 99% cheaper than GPT-5.5 and Claude on output. The advisory brief I wrote six weeks ago is now testable. No single model wins everything. But the door for regulated enterprises is now open.
Read the Article →Subscribers to the Grey Orbits Briefing receive each new article the day it publishes, direct to their inbox or LinkedIn feed.
Long-form analysis on artificial intelligence, cybersecurity, digital currencies, and post-quantum cryptography. Written for boards, investors, and senior leadership. No promotional material.
Prefer LinkedIn?
Subscribe to Briefing on LinkedInBoard engagements, advisory mandates, and education programmes.