Disney Laid Off a 31-Year Employee and Called It a Knowledge Management Upgrade

Disney's marketing and brand division has been shrinking in waves throughout 2026. The programme began in April, under new CEO Josh D'Amaro and the company's first-ever Chief Marketing Officer, Asad Ayaz, and has cut roughly 1,000 roles across publicity teams, the home entertainment division, and brand functions spanning the studios, television networks and ESPN.

Early July reporting on one of the more recent cuts in this wave surfaced a detail that made the story notable beyond its raw scale: a 31-year Disney employee was let go, with the departure explicitly framed - in reporting that followed - as part of a shift toward AI-integrated knowledge management systems intended to replace what had previously existed only in one person's accumulated experience.

What thirty-one years actually contains

It is worth pausing on what three decades inside a single organisation typically represents, because the framing of "knowledge management" tends to flatten a genuinely varied set of institutional assets into a single, system-shaped category.

Some of what a 31-year employee holds is genuinely a documentation and retrieval problem: which vendor was used for a particular campaign in 2011, what a specific brand guideline says about co-branding with external partners, how a past product launch actually performed against its targets. This category of knowledge is, in principle, capturable - it can be written down, indexed, and retrieved by a system as effectively as it could be retrieved by asking the person who remembers it, assuming the underlying documentation exists and was ever properly captured in the first place.

Other parts of that same person's value are not documentation problems at all. Why a particular stakeholder responds well to one style of pitch and poorly to another. Which decision-makers actually hold influence versus which ones merely appear to on an org chart. What went wrong, in ways too political or too subtle to ever be written into a formal post-mortem, the last time a similar campaign was attempted. This second category is experiential and relational in a way that resists indexing, because much of it was never explicit enough to write down even at the time it was learned.

The bet Disney is implicitly making

Every organisation that replaces a long-tenured employee with a system is making an implicit bet about the ratio between these two categories - how much of that person's value was genuinely retrievable information, and how much was tacit, relational, and experience-dependent in a way no system currently captures. Disney's decision, in this specific case, represents a bet that the retrievable-information share is larger than conventional wisdom about institutional knowledge would suggest.

That bet may prove correct in specific, well-documented functional areas - brand guideline compliance, historical campaign performance data, vendor and contract management are all genuinely systematisable. It is far less obviously correct in relationship-dependent functions, where the actual value of a long-tenured person often lies less in what they know and more in who trusts them, and trust is not a retrieval problem.

The pattern beyond Disney

Disney is a highly visible example of a decision playing out quietly across many organisations, including much smaller agencies and businesses that will never generate a headline. Every time a business considers whether a senior, expensive, long-tenured person's role can be absorbed by a system plus a more junior replacement, it is making the same implicit bet Disney just made publicly - and, in the great majority of cases, doing so without the scrutiny a public company's layoff decisions attract.

The risk in getting that bet wrong is not immediately visible. In the short term, dashboards keep working, campaigns keep launching, reports keep generating. The failure mode of over-estimating how systematisable a senior person's knowledge actually was tends to surface much later - a client relationship quietly cools because the person who understood its history and sensitivities is gone; a campaign repeats a mistake that was never written down anywhere because the only record of it existed in someone's memory; a stakeholder disengages because the trust that had been built over years evaporated along with the person who built it.

A more honest way to make the same decision

None of this means long-tenured, senior staff are automatically irreplaceable, or that documentation and knowledge-management systems have no genuine value - they clearly do, for the genuinely systematisable share of what any experienced person holds. The more useful discipline, before making a decision like the one Disney made, is separating the two categories explicitly rather than treating "institutional knowledge" as a single undifferentiated asset that a system either replaces or doesn't.

A practical version of this exercise: for any senior, long-tenured person whose role is being considered for reduction or replacement, list specifically what they know that is written down or writable (documentation, historical data, process knowledge) separately from what they know that exists only because they were personally present for it and personally built the relationships around it (stakeholder trust, undocumented political context, judgment calls that were never formalised into a rule). The first category is a legitimate target for a knowledge-management system. The second category is not, regardless of how sophisticated the system is, because it was never actually a documentation problem to begin with.

Disney's decision - and the many smaller, quieter versions of the same decision happening elsewhere - will be judged not by whether the system it installed is impressive, but by whether the specific 31 years of relational and tacit knowledge that walked out the door turns out, in hindsight, to have been more of the second category than the first.