Stock Markets February 21, 2026

Analysts: AI Will Augment, Not Erase, Software Engineering Jobs for Now

Bernstein says coding automation affects only a slice of engineers' work; market reactions may be overstated

By Avery Klein
Analysts: AI Will Augment, Not Erase, Software Engineering Jobs for Now

Bernstein analysts argue that widespread predictions of AI displacing software engineers are exaggerated. They point to limited real-world adoption, modest productivity improvements, and the fact that coding comprises less than a fifth of an engineer's time. While AI-linked layoffs have been notable, Bernstein views AI as an incremental layer in enterprise technology rather than an immediate substitute for knowledge workers.

Key Points

  • Coding represents less than 20% of a software engineer's work; most time is spent on documentation, deployment, integration and process management. - Sectors impacted: IT services, enterprise software, cloud infrastructure.
  • Nearly 80% of developers use AI tools, but generated code frequently requires debugging and refinement, limiting immediate productivity gains. - Sectors impacted: software development tools, DevOps services.
  • Tech layoffs since 2020 exceeded one million globally; Bernstein links much of this to excess hiring and weaker demand, noting AI-linked sectors have seen a higher share of cuts. - Sectors impacted: technology, AI-focused units within larger firms.

Concerns that artificial intelligence will rapidly supplant software engineers and hollow out the IT services industry are, in the view of Bernstein analysts, overstated. The firm signals that, to date, actual adoption of AI coding tools remains constrained and the productivity gains reported so far are modest.

Bernstein highlights that writing code represents under 20% of a typical software engineer's responsibilities. The remainder of engineers' time is devoted to tasks such as documentation, deployment, systems integration and management of development processes. In that light, even if AI systems can generate code more quickly than humans, they would only automate a minority of the overall job.

Survey data referenced by the analysts indicates that nearly 80% of developers now use AI tools in some form. However, a substantial share of those users report that code produced by the tools still needs refinement and debugging before it can be deployed. That gap between initial generation and production-ready code dampens the immediate disruptive potential of these systems.

Layoffs in the technology sector have remained elevated since the post-COVID hiring surge, with more than one million workers cut worldwide since 2020. Bernstein attributes much of this workforce reduction to over-hiring during the hiring boom and weaker demand conditions. The firm notes, however, that sectors tied to AI have accounted for a higher proportion of recent cuts. Some companies may have trimmed headcount partly to signal AI adoption, yet Bernstein emphasizes that the tools themselves are not yet substituting for teams at scale.

The analysts explicitly reject comparisons between the current AI environment and the late-1990s dot-com bubble. They point out that leading AI companies today are established, profitable firms that are investing from robust cash positions, rather than speculative startups dependent on speculative capital.

Looking ahead, Bernstein expects AI to layer incrementally into enterprise technology stacks rather than to function as an immediate replacement for human knowledge work. "The current market reactions perfectly capture two fallacies - the slippery slope and the bandwagon, and once the dust of the AI storm settles, we perceive a correction to emerge," the analysts write. They add that, while longer-term disruption cannot be ruled out, current market sell-offs in AI-exposed sectors imply an economic shock that is not yet visible in corporate behavior or macro data.

On the investment side, Bernstein suggests that the heightened fear in markets can create buying opportunities. "For the investors, periods of heightened fear offer opportunities to buy beaten down sectors," their note concludes.


This assessment frames AI as a technology that will likely augment enterprise capabilities over time, with immediate effects concentrated in specific coding tasks rather than across the full spectrum of engineering functions. The trajectory and timing of broader disruption remain uncertain, and market pricing appears to have moved faster than underlying corporate actions.

Risks

  • Longer-term disruption from AI cannot be ruled out; while current evidence points to augmentation, extended timelines or new capabilities could change labor demand dynamics. - Affects: labor markets, IT services.
  • Market sell-offs in AI-exposed sectors may reflect expectations of an economic shock that is not yet visible in corporate behavior or macro data, introducing valuation risk for investors. - Affects: equity markets, AI and enterprise software stocks.
  • Some companies may be reducing headcount to signal AI adoption rather than because tools have replaced teams, which could obscure true productivity and operational changes. - Affects: corporate governance, investor sentiment in tech.

More from Stock Markets

NYSE Holdings UK Ltd launches unified trading platform to streamline market access Feb 21, 2026 Earnings Drive Weekly Winners and Losers as Buyout Headlines Lift Masimo Feb 21, 2026 Barclays Sees 'Physical AI' Scaling to Hundreds of Billions by 2035 Feb 21, 2026 Germany's Wind Expansion Accelerates Amid Growing Questions Over Durability Feb 21, 2026 Bernstein outlines framework to assess AI disruption risk across software stocks Feb 21, 2026