Gartner's strategic planning assumption: by 2029, 50% of software application providers will be forced to share their context layer externally for third-party orchestrators to stay relevant. Today, that number is less than 2%.

This isn't a gradual evolution. It's a structural shift in where value accrues in the software stack — away from the systems that hold data and toward the platforms that can contextualize and deliver it to AI agents at the moment of need. Context is what agents eat. Via APIs and MCP tools. Enterprises that don't build the infrastructure to participate in this economy won't just miss an opportunity. They risk being disintermediated by players who can.

Here's the reality most AI strategy conversations miss: enterprises aren't starting from zero. The richest context in any organization already exists — in API traffic, in event streams, in transactional data, in the behavioral signals generated by different systems of record. The question isn't where to find context. It's whether you have the architecture to govern it, orchestrate it, and make it securely available to AI agents across your organization and your partner ecosystem.

Most enterprises don't. They have a fragmented infrastructure built for a world where humans, not agents, were the primary consumers of digital services — divided by different BUs in their own silos. But AI agents have fundamentally different requirements. They need context that is across the entire enterprise, enriched, semantically staged, and delivered in real time across protocols that most enterprise architectures weren't designed to support.

Gartner identifies a real tension here. Enterprises that try to protect their competitive position by locking down their context will find themselves increasingly irrelevant as AI orchestration moves to platforms that can aggregate context across multiple sources but still control the who and how. But enterprises that open up their context without proper governance, security, and monetization infrastructure face a different set of problems: compliance exposure, cost overruns, and value leakage.

The enterprises that win in the context economy will be the ones that thread this needle: opening up their context strategically, with the right controls in place, and with a business model that ensures the value of that context is captured rather than given away.

That requires more than a gateway or an AI tool. It requires a platform that spans the entire context lifecycle — from the APIs, LLMs, and event streams where context originates, through the governance and orchestration layer where it's managed, to the discovery and monetization layer where it becomes a business asset.

In other words, we’re in need of a unified approach to AI connectivity.

At Kong, we've spent years building the infrastructure that governs how data moves across enterprises through APIs, through event streams, and increasingly through AI-native protocols like MCP. 

When I look at Gartner's description of what enterprises need to compete in the context economy, I see a checklist that maps directly to the unified API and AI platform we've built for our customers: unified management across every protocol context travels on, orchestration that prevents fragmentation and lock-in, a catalog as universal system of records for all your composable assets, developer and agent-facing discovery that makes context commercially accessible, and — critically — the metering and billing layer that eventualy turns context delivery into a revenue stream.

That last piece is where I think the market is most underestimating the challenge. Building a context economy strategy without a monetization infrastructure is like building a marketplace without a payment system. You can create supply and demand, but you can't capture value. Kong is the only unified platform that does it all — from system of record to monetization.

The token and context economy isn't a 2029 problem. The infrastructure decisions enterprises make in the next 12 to 18 months will determine whether they enter that world as participants or as bystanders — if they will be enriched or eaten by AI. The good news is that most enterprises already have the raw material — the data, the systems, the context — that the context economy runs on. What they need now is the architecture to unlock it.

That's the AI connectivity problem Kong was built to solve. And if Gartner's analysis is right (and I believe it is), it's the most important infrastructure investment an enterprise can make right now.