The last year has brought a surge of predictions about the impact of Generative AI—on white-collar work, on entire sectors, on business models. Most of these forecasts rest on a simple premise: if GenAI can do a task, the person—or firm—doing that task is at risk.
It’s a clean logic. But it turns out to be flawed.
Over the past 18 months, I’ve led a research project that engaged more than 300 senior executives and directors—via roundtables and surveys with the UK’s Institute of Directors, one-on-one interviews, and AI-based analysis of firms’ market positioning using website and LinkedIn data. We also worked through Evolution Ltd, a boutique advisory on AI, strategy and ecosystems, with private-equity funds and their portfolio companies to understand how GenAI is being interpreted—and misinterpreted—in boardrooms and leadership teams.
What we found was clear: disruption doesn’t start with tasks. It starts with how a firm creates value. And if leaders confuse the two, they risk making the wrong strategic moves.
The Task-Level Fallacy
Much of the discourse around AI, especially in the business press, relies on occupational and task exposure models: if GenAI can write copy, analyze data, or summarize legal documents, then marketing firms, analysts, or lawyers are all at risk. But that approach treats organizations as if they were modular—neatly separable into a sum of tasks.
They’re not.
Even firms whose outputs are task-intensive often rely on value that comes from integration: from how parts fit together, from trust, judgment, and regulatory understanding. What looks automatable at a distance may be deeply entangled with things GenAI can’t yet replicate.
In our work with directors and C-suite leaders, the pattern was consistent. Executives weren’t worried about tasks per se. They were concerned about whether GenAI could credibly challenge their value proposition—what their firm offered that clients paid a premium for.
A legal director told us: “Yes, GenAI can draft contracts. But our value is knowing what that contract means in this case, in this jurisdiction, for this client. That’s not in the training data.”
A strategy head in professional services said: “If I describe our work by tasks, we’re obsolete. If I describe it by outcomes, we’re differentiated.”
Why Modularity Matters
GenAI thrives on modularity. The more separable, repeatable, and pattern-based a task is, the more vulnerable it is to being automated or commoditized. But organizations aren’t simply stacks of tasks—they’re systems. This is what most of the pundits and GenAI enthusiasts miss, conflating massive (and real) productivity gains at the level of the task with impact for organizations – and society.
What matters is how much of your value can be decoupled—and whether your clients still believe the output has strategic meaning when the human is removed.
That’s why some of the firms that looked most “at risk” on paper felt surprisingly confident, while others in supposedly AI-proof industries saw red flags. It came down to how leaders framed their firm’s value—and how much of that value rested in elements that GenAI can replicate or reframe.
The Research Behind the Insight
To understand this more rigorously, we used a three-stage approach. First, exploratory roundtables with UK directors across sectors, identifying how executives thought about GenAI—what excited them, what worried them. Second, a large-scale survey of nearly 300 IoD members to map patterns of perceived opportunity and risk, and test them across different industries, firm types, and leadership roles. Third, follow-up interviews and validation, including a separate roundtable with non-IoD directors, and a novel analysis of firms’ self-described market positioning using NLP tools and archived web data, which also looked at how things have shifted over the last year – when GenAI has advanced by leaps and bounds.
We found consistent patterns:
– Executives who believed GenAI matched their core value proposition reported higher perceived displacement, regardless of task exposure.
– Those who viewed their value as distinct to GenAI’s known strengths felt buffered—even if the underlying tasks seemed automatable.
– Strategic positioning—not job titles or functional maps—best predicted whether leaders were preparing for augmentation or upheaval.
– Even with GenAI’s rapid evolution, most executives we re-interviewed said their views hadn’t changed. Their expectations remained grounded in how they defined their firm’s value—not in which tasks could be automated.
Yet we wondered whether our early insights would hold under the drum‑beat of GenAI headlines and the breathless forecasts of wholesale industry upheaval. So, over April-July 2025 we doubled back: running fresh director roundtables, fielding a new targeted survey, adding one‑to‑one interviews, and scraping hundreds of firm‑website pages with NLP tools to track how companies now describe themselves. The headline? Despite the lightning‑fast technical gains and the louder evangelism surrounding them, the business fundamentals our respondents care about—their core value propositions and the trust‑laden relationships that deliver them—have barely budged.
What This Means for Strategy
If you’re a CEO, investor, or strategy lead, the takeaway is straightforward: your firm’s exposure to GenAI is not about what you do, but about what you believe your value is—and how clearly you can articulate and defend it.
Stop thinking in terms of tasks. Start thinking in terms of differentiation—and how GenAI might enhance it, not just erode it.
Here’s how to get started:
1. Reassess Your Value Proposition
Ask: What do clients or customers pay for that they can’t easily get elsewhere? If the answer is speed, consistency, or output volume—you’re likely in GenAI’s strike zone. If the answer involves trust, domain-specific understanding, or complex interpretation, you have more runway. Use that runway to figure out how GenAI can become a source of strength—not a threat.
2. Diagnose Strategic Modularity
Which parts of your offer can be separated and done just as well (or better) by a system? Which parts require integration, relationship, or deep contextual knowledge? That’s where AI meets strategic risk. Consider redesigning at-risk offerings with GenAI in mind—and build new ways to deliver strategic value.
3. Reframe Exposure as a Choice
GenAI can unbundle your offer. That doesn’t mean it should. Identify where to automate (and be cost-competitive), and where to double down on human-led value (and justify a premium). Use this moment to rethink and rebuild your competitive moats—with clarity about where AI fits and where your edge endures.
A Case in Point
One PE-backed firm we worked with—a mid-sized consulting agency—looked, on the surface, highly automatable. But by mapping its service lines not by task, but by value anchors, the leadership team saw that their stickiness came from how they framed and interpreted information—not just produced it.
Rather than slash headcount, they:
– Augmented delivery with AI tools for speed
– Retooled training to emphasize judgment and framing
– Introduced a new “AI Assurance” layer, billed to clients as part of strategic oversight
The result wasn’t just retained relevance—it was improved margin and better client conversations.
The Risk of Misreading the Moment
The danger isn’t that GenAI will overtake your firm overnight. It’s that you’ll respond to the wrong signals—automating what doesn’t need to be, ignoring what’s shifting, or misjudging what clients actually value.
Leaders who chase AI trends without understanding their firm’s true value risk automating themselves into irrelevance. Those who know what sets them apart—and where AI can support rather than replace it—will come out ahead.
GenAI is moving fast. But strategy still sets the course.
As our work with the IoD and leading firms shows, the firms that survive this wave won’t be the ones that react the fastest. They’ll be the ones that understand what to protect, what to reimagine—and how to tell the difference.
Michael G. Jacobides is the Sir Donald Gordon Professor of Entrepreneurship & Innovation and Professor of Strategy, London Business School and the Lead Advisor of Evolution Ltd. This draws on a project co-authored with M. Dalbert Ma, based on a collaborative research effort with the UK’s Institute of Directors and several private-equity-backed firms, supported by practitioner work through Evolution Ltd. Findings are based on surveys, interviews, AI-driven analysis, and direct senior management and boardroom engagement.