AI Didn’t Kill Engineering

The recent wave of tech layoffs has been widely described as cost cutting or a response to macro conditions. In practice, many of these moves are framed internally and externally as efficiency initiatives, increasingly tied to artificial intelligence.

That pattern is not new. Every major technology shift creates a period where companies promise more output with fewer people. Whatisdifferent this time is where some organizations are making their cuts.

In many cases, the engineers being let go are not the least capable. They are often the most tenured, removed as part of blunt efficiency moves rather than any serious assessment of capability. That is precisely the mistake. These are the engineers who understand systems at scale, who think in architecture rather than prompts. AI does not make that skill set obsolete. It makes it more valuable.

AI tools are now widely adopted across software teams. Code assistants, automated testing, and design-to-code workflows are becoming standard. What those tools reliably do is increase speed. What they donotreliably do is increase judgment.

Used by an inexperienced engineer, AI accelerates the creation of fragile systems. Things get built quickly, but not necessarily correctly. Bugs become harder to trace. Technical debt accumulates invisibly. Used by an experienced engineer, AI becomes a force multiplier. Tasks that once took weeks can be completed in hours, precisely because the person using the tool already understands what a robust system should look like.

This distinction is becoming increasingly relevant in M&A.

Many large organizations are discovering that despite access to AI tooling, their internal teams struggle to deploy it effectively. The limitation is not technology. It’s organizational DNA. Process-heavy environments, fragmented ownership, and shallow engineering experience make it difficult to translate AI capability into real product velocity.

When internal transformation stalls, acquisition becomes the fastest path forward.

We are seeing growing interest in small, highly capable teams that already know how to build and ship in an AI-augmented environment. Not teams defined by headcount, but by capability density. Teams that have built real systems, in production, with limited capital and clear technical ownership.

This is less about “acqui-hire” in the traditional sense and more about acquiring applied competence.

AI is also starting to challenge long-held assumptions about software economics. When experienced teams use these tools well, traditional SaaS cost structures begin to bend. Feature development accelerates. Margins expand. What still protects incumbents is not code alone, but years of embedded data, customer workflows, and operational knowledge. Those assets remain difficult to replicate.

At the same time, the market is quietly recalibrating expectations. AI has been heavily marketed as a solution to everything. Boards and executives are now discovering its limits. It does not fix unclear strategy, poor execution, or weak ownership models.

That disillusionment is not slowing M&A. It is sharpening it.