Closing the AI trust gap: What SitecoreAI DAM + Scrunch mean for your content strategy

By James Gray, Digital Strategy Director, Horizontal Digital

For a while now, we've been telling clients that their digital asset management platform isn't just a backstage operations tool anymore. It's becoming part of how AI represents their brand to the world. Sitecore's latest move is the clearest confirmation yet that the platform itself agrees. 

With SitecoreAI DAM now integrating directly with Scrunch, Sitecore has connected the two halves of a problem that most organizations have been solving separately, if they've been solving it at all: governing what your brand actually says and knowing what AI says on your behalf. 

AI doesn't read your brand deck - it reads whatever it can reach

Before a prospective customer visits your website, downloads a whitepaper, or picks up the phone to your sales team, there's a good chance they've already asked an AI assistant to explain the category, compare their options, and build a shortlist. That assistant isn't drawing on your approved messaging by default. It's drawing on whatever content it can find, structure and trust, and for most organizations that means only a fraction of what's actually sitting in their content estate. 

The result is a gap a lot of brand and marketing leaders don't yet have visibility into: content teams can approve a claim, publish it, govern it carefully through every internal workflow, and an AI system can still hand a customer something incomplete, out of date, or flatly contradictory. Not because the brand did anything wrong, but because governance and AI visibility have, until now, lived in two completely different systems that never talked to each other. This is the gap Sitecore has just built a bridge across. 

Sitecore's answer: turn governed content into verifiable answers

The integration connects three things that used to sit in isolation. SitecoreAI DAM holds the governed version of the truth: the approved assets, claims and content that a brand actually stands behind. Scrunch's Knowledge Studio turns that approved content into structured, traceable Knowledge Documents that AI systems can parse. And Scrunch itself monitors what AI search assistants and answer engines are saying about the brand in the wild. 

What makes this more than another monitoring dashboard is the loop it creates: Govern → Structure → Observe and check → Correct → Verify. Instead of generating another feed of alerts to sift through, every AI claim about the brand gets classified against what's actually been approved, as Supported (matches approved knowledge), Contradicted (conflicts with it), or Not Covered (no approved source exists yet). That last category matters as much as the other two: it tells a brand exactly where its content estate has never given AI anything trustworthy to work with in the first place. 

As Sitecore puts it, the point “isn't to generate more alerts. It's to connect findings to trusted sources, make governed improvements, and determine whether those improvements influence how the brand is represented over time.” That's the accountability piece most AI-visibility conversations have been missing. 

Why this isn't just a marketing-team problem

Governance has always mattered for brand consistency. It matters more now because the cost of getting it wrong has changed. An outdated claim in a shared drive used to be a risk if a person happened to find it. An outdated claim sitting anywhere near your DAM can now be surfaced, summarized and redistributed by an AI system with exactly the same confidence as something current and correct, and the person on the other end has no way of knowing the difference. 

That's why this conversation tends to land differently depending on who's in the room. Marketing and brand leaders hear an opportunity: get this right and you influence how AI represents you before a customer ever reaches your website. IT and architecture leaders hear a governance and integration question: whose system of record is this, how does it connect to what's already in place, and what does “verified” actually mean at platform level. Both are right, and the organizations getting ahead here are the ones building one answer for both audiences rather than pitching this as a marketing initiative that quietly needs IT's permission later. 

Where this leaves organizations already on Sitecore, or weighing it up

For businesses already invested in Sitecore, or moving towards a SitecoreAI ecosystem, this closes a loop that's been conspicuously open. A DAM stops being “the place assets live” and becomes the governed layer the rest of the AI-visibility conversation depends on, which means the decisions that shape it (taxonomy, metadata, ownership, approval workflows) stop being implementation detail and start being strategy. 

It also raises the bar for anyone treating DAM as something to sort out after the bigger platform decisions are made. If Content Hub is going to be the source Scrunch checks AI claims against, the underlying content model needs to be built to support that from the start, not retrofitted once the gaps are already showing up in what AI is saying about the brand. 

How Horizontal helps

As a Sitecore Diamond Partner, this is exactly the territory we spend our time in with clients: assessing what's actually sitting in a content estate today, closing the gap between “governed” and “AI-ready”, and making sure a DAM implementation is designed for where the organisation is heading, not just the pilot in front of it. 

In practice, that means auditing the current content estate against what AI systems can actually find and trust, defining the ownership, metadata and approval model a Content Hub implementation needs to scale, and connecting that governed layer through to Scrunch so brand and IT teams can see (and act on) exactly where AI is getting the brand right, and where it isn't. 

If you want to know where your own content estate stands against this, get in touch with us.