Organizations are complicated techniques composed of people that come and go, very like organic organisms are made up of particular person cells that die and regenerate. However in contrast to organic techniques, the place every cell carries the organism’s full genetic code, institutional information in organizations is fragmented: No single particular person has the whole image. Because of this, organizations invent mechanisms to seize, retain, and transmit information, making certain continuity, at the same time as people cycle out and in. Lew Platt, former CEO of Hewlett-Packard, was proper to lament: “If only HP knew what HP knows, we’d be three times more productive.”
It’s no shock then that the “knowledge management” market is projected to achieve $2.5 trillion by 2030. This explosion in potential worth is due, partially, to the potential and progress of generative AI as a device to excavate buried knowledge and tacit information, recontextualize it, and provide organizations not only a higher understanding of their previous, however a extra coherent strategic imaginative and prescient of their future.
At this time’s commonplace approaches to information administration in organizations are constrained by two limitations. First, they usually fail to seize what might find yourself mattering most: The huge repositories of tacit information that people seldom translate into neat, “final” paperwork. The reasoning behind key choices, the context of near-misses and previous actions, and the instincts that formed an establishment’s tradition—these are the sorts of insights that quietly disappear, eroding a company’s skill to be taught from itself. Second, conventional techniques are constructed on an assumption that organizations can anticipate what information can be wanted or helpful sooner or later.
A latest creative venture led by our co-author, Zoé Vayssières, with Harvard College’s Digital Information Design Institute on the usage of GenAI to uncover and inscribe Harvard’s early “memories” (based mostly solely on uncooked knowledge, like wage data, not present historic accounts) offers a robust illustration of how know-how can’t solely resurface institutional intelligence, however recontextualize it—making reminiscence dynamic, discoverable, and usable in unpredictable, various contexts.
This shift from static information administration to dynamic “memory management” unlocks new potentialities for complicated organizations. Somewhat than counting on predefined classes, GenAI allows organizations to categorise info on the level of retrieval, surfacing related perception based mostly on the questions being requested, not the assumptions made on the time of storage. That responsiveness makes reminiscence not simply simpler to entry, however extra actionable. This isn’t merely a technological shift, however a strategic one: For the primary time, organizations can deal with reminiscence not simply as one thing to protect, however a useful resource to activate—turning gathered expertise into aggressive benefit.
What reminiscence is—and why organizations lose it
Organizational reminiscence is commonly mistaken for record-keeping. However true reminiscence is not only an archive of the previous—it’s the mechanism by which an establishment features coherence, pace, and resilience. It captures the values that influenced management and the drivers of important choices, and shortens the educational curve for brand spanking new expertise. With out it, corporations can undergo “strategic amnesia”: repeating previous errors and drifting culturally with each management change.
And but, for all its significance, reminiscence stays fragile. In contrast to particular person people, organizations don’t bear in mind “centrally”: Their reminiscences should be constructed based mostly on information scattered throughout electronic mail threads, retired techniques, untagged folders, and the minds of people that depart with out being requested what they know. What will get recalled in a company is commonly formed by the construction of techniques (e.g., efficiency metrics) constructed for documentation, not for understanding. On the similar time, the softer types of perception have a tendency to fade.
Paradoxically, the extra knowledge fashionable organizations generate, the harder it has change into to retrieve significant perception: Corporations are flooded with info but starved for readability. The true problem just isn’t lack of knowledge, however the absence of context and the problem of purpose-specific recollection.
GenAI makes higher institutional reminiscence doable
Generative AI adjustments what is feasible by unlocking what already exists, enabling organizations to work together with their unstructured archives at scale and in the end bear in mind extra—unearthing forgotten perception, reconnecting scattered info, and exposing buried patterns.
On this sense, GenAI for reminiscence administration features much less like an organizational historian and extra like a customized company archaeologist: not recording one thing new from scratch however excavating what was already there. In doing this restoration and recontextualization, GenAI can present a deeper understanding of the context of the related details, going again to the “origin” of the information, or a choice, whereas additionally leaving it free to be reinterpreted sooner or later, based mostly on the brand new context of that point. Within the creative venture involving Harvard’s archives, GenAI fashions have been used to research over 200 years of institutional data—a few of them handwritten, inconsistent, or solely partially digitized. The fashions surfaced neglected institutional guidelines, social roles, and emotional registers that might have been difficult for even historians to uncover.
For instance, an evaluation of uncooked historic wage knowledge revealed how organizational priorities can shift each instantly and progressively. In 1752, Harvard’s president earned almost two occasions greater than the steward, charged with managing residential operations. However by 1779, amid wartime upheaval and hovering inflation, the steward’s wage had greater than tripled whereas the president’s had declined—reflecting a shift in priorities throughout disaster.
Throughout a broader timespan, wage patterns additionally revealed the evolving worth positioned on totally different educational disciplines, with the sciences progressively overtaking theology in compensation—signaling a deeper transformation within the establishment’s mission. For enterprise leaders, these patterns provide a robust reminder: In durations of transformation, it’s usually the casual shifts—quite than formal declarations—that reveal the place affect resides, particularly at first. GenAI can floor these hidden dynamics, serving to organizations perceive how they’ve traditionally tailored below stress, and the place tradition and energy might diverge from construction.
One other instance: The resurfacing of ladies whose roles have been instrumental in Harvard’s early improvement however are largely absent from conventional historic accounts. These contributions weren’t misplaced, quite they have been traditionally deprioritized as a result of they didn’t mirror what earlier generations sought to spotlight. However when prompted with right now’s context, GenAI surfaced wealthy element about greater than 10 key girls (together with Squaw Sachem, Anne Dudley Bradstreet, and Elizabeth Glover Dunster) whose affect had lengthy gone unrecognized. The GenAI-enabled retrieval of those contributions challenges assumptions about who holds important information and what roles they play. In company settings, the identical blind spots exist. GenAI allows organizations to revisit their very own histories with contemporary eyes and up to date understanding, surfacing undervalued people, neglected features, or casual techniques that quietly sustained efficiency.
Reminiscence as a strategic use case for GenAI
Whereas a lot of the eye round GenAI right now has centered on automation and productiveness, it has important long-term worth for operationalizing institutional reminiscence as effectively. GenAI can energy onboarding, technique improvement, danger administration, and extra, with wide-ranging purposes: surfacing neglected experiments that will now make sense below new market situations, connecting fragmented threads between enterprise items, and figuring out repeated however undocumented workarounds. In management transitions or onboarding, it might probably protect casual know-how—not simply how techniques work, however how issues get carried out.
A telling instance comes from the French electrical distribution firm Rexel, the place an AI-powered gross sales device designed to suggest next-best presents shortly grew to become greater than only a productiveness enhancer. Newer workers relied on it for steerage, whereas skilled distributors used it to coach the system—confirming or refining its strategies. In doing so, every particular person embedded their tacit information right into a device that might help others. Over time, the platform developed from a advice engine right into a reminiscence system, capturing experience that may in any other case have walked out the door.
An analogous dynamic is taking part in out in software program improvement, the place the AI agent CODAY—developed by Whoz—captures engineers’ tacit, undocumented problem-solving methods and coding kinds. As Whoz CEO Jean-Philippe Couturier explains, “By turning individual intuition into collective, reusable knowledge, CODAY transforms the most intangible aspects of expertise into organizational memory.” Put one other manner, the AI agent doesn’t simply get better perception—it scales it.
When GenAI is used adroitly, organizations are capable of reconstruct, reinforce, and venture ahead their core DNA. It might floor recurring instincts in decision-making, constant tones in communication, and embedded values which have formed the group over time. By making legacy knowledge accessible, with out idealizing it, GenAI can flip reminiscence from a hard and fast repository right into a dynamic, strategic asset: a platform for future motion that helps organizations transfer ahead with better coherence and confidence.
A basis for quicker, smarter change
In durations of speedy transformation, organizations don’t simply want to vary—they should change coherently. That coherence ensures that change just isn’t solely executed, however absorbed, aligned with the establishment’s underlying tradition, values, and strategic intent. As BCG analysis has proven, transformations ceaselessly falter after they fail to account for the casual norms and shared assumptions that govern how work is actually carried out. On this context, reminiscence turns into important. It’s the connective tissue that allows choices to be quick, but aligned with long-term values—decreasing duplication, confusion, and drift. With out it, corporations are left reacting in fragmented methods, dropping sight not simply of the place they’ve been, however of who they’re.
For leaders navigating change, that is each a robust asset and a shift in context: the CEO is now not the only steward of the previous or the lone narrator of the long run. With institutional reminiscence extra accessible throughout the enterprise, management turns into much less about carrying the complete weight of continuity, and extra about drawing from a shared supply of perception. On this context, choices acquire traction not as a result of they’re imposed, however as a result of they resonate with what the group already is aware of and remembers. When workers perceive the reasoning behind prior choices, and the way their group has developed, they have interaction not simply with their function, however with the establishment itself. Strategic reminiscence turns into a device for organizational cohesion, decreasing informational asymmetries, fostering belief, and creating the situations for extra assured, aligned motion at each stage.
The worth of GenAI as a “corporate archeologist” just isn’t about wanting backward. It’s about constructing strategic reminiscence as infrastructure—no much less important than a CRM or knowledge warehouse—which requires embedding reminiscence into onboarding, management improvement, and decision-making. In an age of fixed reinvention, reminiscence is what ensures every reinvention is coherent. Strategic reminiscence, subsequently, isn’t only a document of the previous, it’s an funding within the high quality of future choices, a important supply of aggressive benefit.
Learn different Fortune columns by François Candelon.
Francois Candelon is a accomplice at non-public fairness agency Seven2 and the previous world director of the BCG Henderson Institute.
Zoé Vayssières is an artist and government fellow at Harvard’s Digital Information Design Institute.
David Zuluaga Martínez is a senior director on the BCG Henderson Institute.
Amartya Das is a principal at BCG and an envoy on the BCG Henderson Institute
Among the corporations talked about on this column are previous or current shoppers of the authors’ employers.
This story was initially featured on Fortune.com
