When AI is embedded in your Workforce, Trade Secrets Become the Strategy
If AI can do the work, what exactly does your company still own? That is the question sitting underneath McKinsey’s new report, The symbiotic enterprise, which argues that AI agents and intelligent robots are becoming a workforce — close to 60 per cent of work hours are now theoretically automatable — and that the traditional moats of expertise, scale and coordination erode as a result. Read it with an IP strategy lens and the most important finding is buried in the competitive analysis: as access to frontier models commoditises, durable competitive advantage shifts to proprietary intelligence — unique data assets, “agentic skills” that encode how your organisation actually works, and learning loops fed by your own operations. Every one of those is an intangible asset, and most of them are protectable, if at all, as trade secrets and contractual rights. Consider the report’s own example: a claims-handling skill encoding your escalation rules, pricing thresholds and compliance logic. That is decades of operational know-how, extracted from people’s heads and written down in deployable form — enormously valuable, trivially copyable, and only defensible if you have deliberately made it so.
The strategic work, then, is to treat the intelligence layer as an IP portfolio from day one. Three questions to consider now. First, ownership: when know-how is codified into agent skills — often with a vendor’s platform and people involved — do your contracts actually vest those skills, and the improvements from learning loops, in you? That engagement-and-ownership problem is the subject of AI Transformation Is Not a Strategy Problem — It’s an Ownership Problem. Second, secrecy: trade secret protection survives only with reasonable steps — access controls, provenance logging, and limits on what agents (and their providers) can see and disclose — which is precisely where autonomous systems create new leak paths, as explored in Your AI Agent Won’t Keep a Secret. Third, stewardship: compounding assets decay without active management, the pattern examined in The Quiet Decay: Why IP Value Slips When No One Is Watching. McKinsey warns of a “cognitive tax” — value flowing to the AI providers everyone depends on. The companies that avoid paying it twice will be those whose data, skills and know-how are identified, owned and protected as rigorously as any patent family. In the symbiotic enterprise, IP strategy isn’t just a legal workstream within the AI programme. It is the part that decides who keeps the value.

