Rocket Software, whose data modernisation business helps large enterprises manage decades of accumulated operational data, argues the deal is less about Spirit Airlines specifically and more about a structural shift in how AI models are built and deployed. As models specialise further, access to proprietary business data becomes competitively valuable rather than merely useful.
The argument cuts in two directions. More industry-specific training data can make AI agents considerably better at the domain-specific workflows they are asked to run. But it also creates a risk of disintermediation: as the example of Anthropic's Claude Legal plug-in illustrated — Thomson Reuters stock fell 16 percent and Wolters Kluwer dropped 13 percent in its wake — frontier model providers gaining access to professional knowledge can undercut the specialist vendors who had previously packaged it.
For most large enterprises, the genuinely sensitive data is not publicly available. It lives inside the organisation: embedded in operational histories, internal applications, workflows, and proprietary knowledge built up over decades. Deals of this kind, according to Rocket Software's president of data modernisation Michael Curry, give frontier models a path deeper into those business processes. Which, in turn, explains the growing enterprise interest in sovereign data and sovereign AI — approaches that allow organisations to adopt AI capabilities without ceding control over the underlying knowledge that differentiates their businesses.
The agentic layer adds another dimension. As AI agents become capable of executing actual business processes rather than merely answering questions about them, the knowledge captured in how those agents operate starts to become as valuable as the source data they draw on. Where that knowledge ends up, and whether it flows back to frontier model providers, is a question many enterprise architects have not yet been asked to answer.
To stay across the latest in cloud, AI and enterprise tech analysis from Compare the Cloud, subscribe to our weekly newsletter at https://www.comparethecloud.net/newsletter
