Something has shifted in the language of corporate restructuring. Where CEOs once cited “market headwinds” or “strategic realignment,” the phrase of the moment is “AI transformation.” The layoffs look the same. The euphemism is just newer.
In 2025 and 2026, more than 180,000 tech jobs were attributed to AI-driven automation, roughly 489 per day. Oracle cut up to 30,000 roles. Amazon eliminated 16,000 corporate positions. Accenture, McKinsey, Atlassian, and dozens of others followed with announcements framing workforce reductions as the inevitable consequence of an AI era. Some of those decisions were genuinely strategic. Others were not.
The problem is that when trillion-dollar companies use “AI is changing everything” as cover for decisions that would have happened anyway, they create a blueprint that every CEO in the country feels pressure to follow. That pressure is worth examining carefully, especially for leaders of middle-market businesses who have far less room for error. And it reinforces the undercurrent that “AI is the bad guy”.
When AI Becomes a Convenient Explanation
Research from MIT and Oxford found that 95 percent of companies investing in AI are generating zero return on that investment, yet AI attribution in layoff announcements continues to rise. That gap between narrative and reality should give every leader pause. It suggests that in many cases, AI is being used not as a genuine driver of workforce decisions, but as a socially acceptable explanation for them.
This matters for two reasons. First, it obscures what’s actually happening, making it harder for employees, investors, and the public to understand the real logic behind cuts. Second, it sets a false standard, implying that any company not aggressively restructuring under the AI banner is somehow behind. Neither consequence serves businesses or the people who work in them.
That said, AI genuinely is changing how work gets done. The honest version of this conversation acknowledges both things at once: yes, AI will reshape roles and workflows in meaningful ways over the next decade, and no, that does not mean the right response is to move fast and break your workforce now.
The Middle-Market Difference
Large enterprises can absorb a strategic mistake in ways that smaller businesses cannot. Meta posted a record $56 billion quarter while simultaneously planning $145 billion in AI capital expenditure. When a company of that scale miscalibrates, it has the financial depth, the headcount, and the institutional inertia to course-correct. Middle-market businesses, typically defined as those with $10 million to $1 billion in annual revenue, do not have that cushion.
What they do have is something more valuable in this moment: the ability to be deliberate. A mid-sized manufacturer, distributor, or services firm doesn’t have ten layers of management between a decision and its consequences. That closeness to the work is an advantage, if leaders use it intentionally rather than treat it as a liability that forces them to move faster to keep up.
The companies that will navigate this era best are not the ones moving quickest. They are the ones moving the most thoughtfully, with careful intention given to critical decisions around their team composition and size.
Plan Deliberately
The first principle for middle-market leaders is to resist the instinct to treat AI strategy and workforce strategy as separate conversations. It’s the same conversation. Any decision about deploying AI in a business process is simultaneously a decision about the people who currently own that process and it deserves the same rigor.
That means starting with a clear-eyed audit: where are the genuine inefficiencies AI can address, and where is AI being considered because it feels like the thing you’re supposed to do? The distinction matters enormously. AI applied to a real bottleneck (a quoting process that takes three days, a customer inquiry queue that’s always behind) can transform a business. AI deployed because a competitor announced something in a press release is just noise with a capital budget attached.
Leaders should also be honest about where they are in their data and infrastructure maturity. AI tools are only as good as the data they operate on. A business that hasn’t yet built clean, integrated data systems will not get the outcomes it expects from AI deployment, regardless of how sophisticated the model is.
Pilot Before You Scale
The second principle is sequencing. The instinct under competitive pressure is to deploy broadly and quickly. The smarter move, particularly for organizations without large AI or data science teams, is to run contained pilots before committing to enterprise-wide rollouts.
A pilot is not a hedge or a sign of timidity. It is the fastest way to understand what actually happens when an AI tool meets the real complexity of your workflows, your customers, and your edge cases. Models that perform beautifully in demos frequently behave differently in production environments where data is messy, context is ambiguous, and the stakes are real. The cost of discovering that at scale is dramatically higher than discovering it in a contained test.
Effective pilots have a defined scope, clear success metrics, and a specific human owner who is accountable for observing what the AI gets right and what it gets wrong. That last point is critical. We are still early. Today’s models make confident mistakes.
They can produce outputs that are fluent, plausible, and wrong. A human in the loop is not a brake on AI capability. It is the mechanism by which you learn what to trust and what to verify, and it is what makes it possible to scale responsibly. It is also most critical for middle market businesses to keep humans in the loop on core processes where a mistake could take a company down (e.g. demand forecasting, inventory purchasing, website updates).
Redeploy Before You Displace
The third principle is about people, specifically, the sequence in which workforce decisions get made. The most common failure mode in AI restructuring is treating displacement as the first move rather than the last resort.
Before any role is eliminated because AI can handle a function, the better question is: what can the person in that role do that AI cannot? AI is genuinely good at high-volume, pattern-based tasks including processing, routing, drafting, summarizing. It is not good at judgment under ambiguity, relationship-building, contextual problem-solving, or the kind of institutional knowledge that lives in people who have been doing a job for years. Most roles contain both kinds of work. The opportunity is to strip out the former and refocus people on the latter, not to eliminate the role entirely because one component of it became automatable.
This redeployment logic isn’t just the ethical approach. It’s also the strategically sound one. The businesses that come out of this period with the deepest competitive advantage will be the ones that figured out how to make their people more effective with AI, not the ones that replaced the most headcount the fastest.
Don’t Let Someone Else’s Press Release Dictate Your Strategy
The final principle is perhaps the most important, and the hardest to hold onto when the noise is loud. Your workforce strategy should be derived from your business, not from the headlines.
When a major company announces AI-driven cuts, it creates a ripple effect. Boards ask questions, investors make comparisons, peers wonder if they’re moving too slowly. That pressure is real, but it is not a strategy. Every business has a different cost structure, talent base, customer relationship model, and competitive context. The right AI roadmap for a 500-person distribution company looks nothing like the right roadmap for a global consulting firm, even if both are reading the same news.
The leaders who will regret this period most are those who made irreversible decisions about their people in response to someone else’s announcement. The leaders who will look back on it best are the ones who stayed grounded in their own business context, moved at a pace their organization could actually absorb, and treated their workforce as the asset AI is meant to strengthen, not the line item it replaces.
AI is the most consequential tool of our generation. That is precisely why it deserves more deliberate stewardship, not less.
