A Call to Arms for Information Professionals

Apparently, the AI revolution is being led by visionary data scientists, brilliant and enigmatic AI ethicists, and the ever-present cadre of advanced software engineers. Billions in investment flow their way, and the narrative is set. But as an industry analyst who spends his days examining these so-called transformations, I can tell you a story of a profession being willfully ignored, its expertise sidelined in a reckless and potentially catastrophic land grab for data. The stark truth is this: There has never been a time in recent history when the professional skills of knowledge and information managers (KIM) have been more desperately needed, yet so conspicuously absent from the conversation.

Look beyond the financial bubble and the hype cycles. On the ground, a truly vast and messy operation is underway—the forced transformation of unstructured, often chaotic, corporate information into “AI-ready” data. Long-locked vaults and silos—containing decades of documents, emails, presentations, and legacy knowledge—are being cracked open and pillaged with abandon. In the frenzied quest to amass training data and fuel models, foundational issues are being trampled. Personally Identifiable Information (PII), security protocols, access controls, and regulatory compliance are not just afterthoughts; they are often viewed as inconvenient speed bumps on the road to AI glory.

This isn’t speculation. In our advisory work at Deep Analysis, we regularly encounter well-funded AI startups and even established large tech vendors barreling headlong into this morass. The most telling detail? Many of these organizations know they have a profound information governance problem. They are acutely aware of the importance of information architecture, taxonomy, and lifecycle management. Yet, they go out of their way to ensure that they have no formal association with our profession. Why? The answer is as blunt as it is absurd: It’s not considered “cool.” In the current climate, any association with what is perceived as a legacy, back-office function is seen as detrimental to securing the next round of venture capital funding or closing a marquee deal. It is more advantageous to label the work as “data curation” or “prompt engineering” than to admit that you need a seasoned knowledge manager.

Thus, we face a profound and frustrating dilemma. Our profession understands, with crystal clarity, that the future of KM and enterprise AI is not just unworkable without us—it is dangerously unstable. AI built on rotten information foundations will fail and fail expensively. It will produce hallucinations rooted in corrupted data, violate regulations with impunity, and leak secrets at scale. Yet, the AI industrial complex either doesn’t care or, if we are being generous, suffers from a catastrophic blind spot regarding our critical role in the equation.

Engaging in Guerrilla Warfare

This leaves us with a choice. We can wait patiently for the inevitable, costly failures to mount to a level that forces a recalibration—a strategy of managed decline. Or, we can engage in some intelligent guerrilla warfare. Here’s how KM and information management professionals can start fighting back: 

1. Infiltrate, Don’t Confront. Stop trying to own “AI” as a project. Instead, embed yourself into every AI initiative as the essential enabler. Volunteer for the “boring” work: auditing training datasets, mapping data lineage for the compliance team, or cleaning the prompt libraries. This isn’t grunt work; it’s a Trojan horse. Once inside, you expose the risks and complexities that others are glossing over. You become the person who asks, “Where did this data come from?” “Do we have the rights to use it?” “What bias does this source introduce?”