As I prepare for DAM New York next week, one question is shaping what I want to learn: What does your organization know about its content that your DAM does not?
Most mature DAM environments contain more than one library. The first is visible: product imagery, campaign creative, video, brand assets and the metadata used to organize them. The second is largely invisible. It is the knowledge surrounding those assets that never made it into the system.
A hero image may be approved for Germany but not for paid social. A pack shot may predate a reformulation. An asset may carry an approved status even though the brand team knows it needs another review before reuse. That context often lives in inboxes, spreadsheets, chat threads and the memories of experienced colleagues.
For years, organizations have worked around this gap because people can interpret ambiguity. They notice when something looks wrong, recognize a familiar exception or ask a colleague for clarification. AI agents cannot read between the lines. They can act only on the context, rules and permissions available to them.
DAM New York's 2026 theme, The Intelligent Evolution of DAM, reflects a shift from static repositories toward systems that can automate workflows, enrich metadata and support decisions. That direction is exciting, but it also makes familiar problems more consequential.
If rights data is incomplete, taxonomies are inconsistent or the real approval process exists outside the platform, adding AI does not resolve the weakness. It allows the weakness to move faster and farther across the content supply chain.
That is why having AI features is not the same as being AI-ready. Readiness depends on whether the DAM represents the way the organization actually works. A field labeled approved may need to answer several additional questions: approved for which channel, geography and time period? Under what license? Can the asset be modified? Can it be used by a generative AI system? Does a person still need to review it?
These are not entirely new governance questions. What has changed is that the answers increasingly need to exist as structured, dependable data—not institutional memory.
Jarrod Gingras's session on the journey to DAM 4.0 positions DAM as an intelligent responder: a platform for orchestration and real-time decision-making, not simply a place to store and find files. I want to hear how organizations are sequencing that journey when their current environment includes legacy technology, fragmented integrations and years of accumulated process.
The destination matters, but so does the route. Which foundations create value now? What must be fixed before an organization introduces more automation? And where can teams make progress without waiting for a perfect data model or a complete platform replacement?
The agenda's focus on business archaeology gets at a reality familiar to anyone who has worked through a DAM migration or transformation. The workflow represented in the system is not always the workflow people follow. Teams change, workarounds become routine and documentation falls behind.
Before automating a process, organizations need to discover the operational truth: who makes each decision, which exceptions matter, where rights and approvals are recorded, and which knowledge disappears when an experienced colleague leaves. AI may help uncover patterns, but it can also amplify a flawed or misunderstood process.
SharkNinja's session offers a useful test case. Its published agenda describes a DAM that grew from 52,000 to 261,000 assets and from 527 to 2,800 global users. With a two-person team managing 2,500 to 3,000 uploads each week, AI-assisted metadata is a practical response to scale.
The question I want to explore is not only what the team automated, but what it chose to keep human. As intelligent tools take on repetitive work, DAM leaders need a clear framework for the decisions that require expertise, context or accountability. Human review should be intentional, not simply the residue left behind after automation.
Sessions on metadata, authenticity, provenance and governance point to another important shift. When content moves from a DAM into creative tools, commerce platforms, partner ecosystems or AI-driven experiences, the governing context must move with it.
A machine knowing that it can access an asset is not the same as knowing whether it should use it. The first is a permissions question. The second is a governance question, and it depends on reliable information about origin, ownership, consent, modification, market restrictions and intended use.
When a person encounters incomplete metadata, the outcome may be a slower search or a message to the brand team. When an agent encounters it, the agent may make the wrong decision, or no decision at all. Metadata quality therefore becomes more than a findability issue. It becomes a condition for safe action.
The organizations best prepared for intelligent DAM may not be the ones with the longest list of AI features. They may be the ones that have done the harder work: clarifying what their content means, making workflows match reality, establishing ownership and deciding where human judgment belongs.
That work sits across people, process and technology. It is also where DAM creates value beyond the platform itself, connecting content to the operating model that enables teams to use it confidently at scale.
At DAM New York, I will be listening for practical examples of how leaders are closing the gap between ambition and operational readiness. If you will be there too, I would welcome a conversation about the knowledge your DAM captures today, the context still living outside it and what needs to change before agents become everyday users of enterprise content.