DAM managers are custodians of context – not just metadata

I recently wrote for DAM News about a belief I think is becoming more common: that AI will replace DAM managers because it can create metadata faster than a person can. It is certainly true that AI can describe images, suggest keywords and process huge collections at speed; those are useful capabilities, and I think they have an important place in the future of digital asset management.

But I do not think that makes DAM management self-governing.

I was pleased to see DAM News’ commentary on the article return to the point that matters most to me: the difference between AI identifying an asset and a human understanding that same asset, and deciding whether or not to authorise its use. It’s an easy distinction to overlook, but it sits at the heart of trusted digital asset management.

I’ve seen DAM managers, DAM librarians, image library coordinators, whatever their job title, spend far more time than people realise on the information around an asset – its rights, permissions, consent, access, relevance and history. That context is often invisible to someone looking at a thumbnail or a metadata field, but that depth of understanding is exactly what allows a marketing team to use content confidently.

That’s why I see DAM managers as custodians of context, and far more than metadata specialists.

Metadata is only part of the DAM manager’s role

Metadata matters. Of course it does! It makes assets easier to find, helps people understand what they’re looking at, and it gives a collection useful structure. Automation can take a considerable amount of the repetitive work out of describing and organising content.

However, in my experience, a well-managed DAM relies on work that is much harder to see. DAM managers are often the people who help decide:

• who should be able to access sensitive content;
• which photographer’s licence applies to an asset;
• whether consent covers a proposed campaign or channel;
• when content should be restricted, archived or removed;
• how external agencies and partners should use content; and
• which information users need before they download a file.

These decisions rely on organisational knowledge. The answer may sit in a consent release form, a contract, brand guidance, a policy document or a conversation with the team that commissioned the work. It may also change over time.

The reason I’ve focused on this is simple: without that context, a DAM can be a well-labelled collection of files, but it is not necessarily a trusted service.

Identifying an asset is not the same as authorising its use

This is the distinction I think deserves more attention as AI becomes more capable.

An AI tool may recognise a group of students, diners, guests or customers in a photograph. It may give a very convincing description of the scene. But it cannot reliably infer the terms that govern how that image may be used.

I have seen how many separate questions can sit behind what looks like a straightforward image. Do the people shown have the right consent? Does that consent cover paid social advertising as well as internal communications? Is the photographer’s licence valid for this territory? Should an external agency be able to download the asset? Is there a safeguarding consideration? Is the image still representative of the organisation?

Those are not metadata questions alone. They are questions of usage rights, consent, permissions and judgement.

The software can apply a rule. But someone needs to decide what the rule should be, which exceptions matter and when that rule needs to change. I think that is one of the clearest examples of the value a DAM manager brings.

For marketing teams, it is also central to staying compliant – and not even having to worry about compliance because it’s all in-hand! The aim is to make sure the right rights and consent information is attached to each asset and easy to see when people need it, so they can get on with using content confidently.

I think automation should improve DAM, not dilute accountability

I am not arguing against automation; I think AI-powered metadata and workflow automation can improve the speed and consistency of DAM. They can help teams ingest content, suggest descriptions, connect records of consent, and apply established access controls.

What I am arguing is that automation should not remove accountability.

Manual metadata errors obviously still occur, and no DAM manager would claim that people are infallible. The difference with automation is scale. One unsuitable keyword may affect a handful of assets, but, when AI is in charge, that same mistaken assumption can be applied to thousands of assets before anyone notices.

That’s why I believe human oversight has to be meaningful. It cannot be a rushed approval step added after a large automated process has finished. DAM managers need the time and authority to set the guardrails, review the outputs and correct problems at their source.

In practice, that means DAM professionals can focus more of their time on work such as:

• designing taxonomies and controlled vocabularies that reflect how the organisation works;
• establishing authoritative sources for rights, consent and asset information;
• setting confidence thresholds and escalation paths for automated metadata;
• sampling automated outputs and identifying systematic errors;
• recognising outdated language, bias and missing organisational context; and
• maintaining an audit trail for decisions that affect asset use.

I believe that this is where automation is most valuable: it can reduce repetitive work, while allowing people to concentrate on governance and quality.

Rights and consent management need context

Rights and consent ‘rules’ are rarely straightforward.

A licence may limit use by channel, territory, audience, campaign, expiry date or partner. Consent for use might be withdrawn, renewed or valid only for a particular purpose. An image might be suitable for an internal presentation but not for a public recruitment campaign.

That’s why I believe usage rights and consent management should be part of everyday DAM workflows. When the relevant conditions sit alongside the asset, teams can make better decisions before a file is downloaded or published.

At Asset Bank, we believe the best digital asset management is the best way to stay compliant: helping teams use assets confidently, control access and protect rights. For me, that means making permissions, usage rights and consent practical for the people who use content every day – not expecting marketers to search through disconnected records before every campaign.

DAM managers help organisations remember what matters

The context around an asset also matters when it is time to decide what to keep.

I’ve certainly seen content that is no longer right for current marketing but still has real historical value. The opposite can be true too: a newer image can become unsuitable quickly because it features former leadership, outdated branding, an old product or a location that no longer represents the organisation.

An age-based retention rule cannot make those distinctions on its own. A DAM manager understands why an asset was created, how it has been used and what its future value might be. They can help decide whether it should remain available, be restricted, moved to an archive or removed from day-to-day use.

The reason I've included this in the conversation about AI is that preserving files is not the same as preserving knowledge. If the people who hold this context are overlooked, an organisation may retain the assets while losing the reasoning that makes the collection meaningful.

Adoption and change management are DAM work too

The most capable DAM cannot deliver value if people do not understand it or trust it. DAM librarians make the system work in practice. They onboard users, explain upload standards, support colleagues and adapt workflows based on real behaviour.

They also connect teams with different perspectives on content: marketing, design, communications, archives, legal, IT, agencies and regional offices. In my view, translating policy and technical features into clear, workable guidance is a core governance responsibility.

This becomes particularly important when organisations introduce AI features or new integrations. Someone needs to ask whether the tool solves a genuine problem, works with existing governance and creates an acceptable level of risk. I feel that DAM managers are often best placed to make that assessment.

The future of DAM is human expertise, amplified by AI

It’s my opinion that the role of the DAM manager will evolve. Repetitive work should be reduced where technology can do it accurately and safely, and that should create more capacity for the work that needs informed judgement: taxonomy design, rights and consent management, permissions, quality assurance, training and collection strategy.

But I do not think the need for stewardship will disappear. If an organisation automates the visible tasks while overlooking the less visible responsibilities, it does not create a self-managing DAM – it creates an under-managed one.

A well-run digital asset management system is not just a database of accurately labelled files. It is a trusted service that helps people find the right content, understand how it may be used and keep it in the right hands. It bolsters confidence. It supports disparate teams in cohesive working.

AI can process a collection. I think DAM professionals are the people who make it useful, trustworthy and accountable.

If you want to chat more about how Asset Bank helps marketing teams manage usage rights, consent and permissions in one trusted DAM, drop us a message or book a demo with the team.

Frequently asked questions

Can AI replace a DAM manager?

I do not think so. AI can accelerate image description, keyword suggestions and bulk processing, but a DAM manager is still needed to set governance, interpret usage rights and consent, decide access policies, review exceptions and keep automated workflows accurate and appropriate.

What is the difference between image recognition and usage rights management?

Image recognition identifies what an asset depicts. Usage rights management records whether, where, how and for how long the asset may be used. An image can be accurately recognised while still being restricted by a photographer’s licence, model consent, location permission or campaign-specific condition.

Why is consent management important in a DAM?

Consent management helps marketing teams understand whether people shown in an asset have permission for the intended use. Linking consent records and restrictions to assets helps prevent avoidable misuse, protects individuals and supports compliant content use.

How can a DAM improve content compliance?

A DAM, or image library, can improve compliance by making rights, consent and access controls part of the asset workflow. Teams can see restrictions before download, apply permissions to sensitive collections, manage expiry dates and use content with greater confidence.

What should DAM managers oversee when using AI metadata?

I think DAM managers should oversee taxonomy design, data sources, confidence thresholds, sampling and quality assurance. They should also review exceptions, check for outdated or unsuitable terminology, correct systematic errors and keep an audit trail for automated decisions.

Further reading

You can read my original DAM News feature and DAM News’ commentary on the article.