Key Takeaways
- Nearly one‑third of organisations (32 %) have opted to forgo buying a software product or feature because they can build it internally with AI‑powered coding agents.
- The “build‑instead‑buy” trend is strongest in technology and healthcare, followed by professional services and energy/materials.
- Forty percent of companies with over $1 billion in annual revenue are now scaling AI agents in at least one function, up from 27 % a year earlier.
- AI high‑performing firms are twice as likely to scale coding agents and 2.7 times more likely to expand other agentic AI use compared with peers.
- About one in five organisations cite AI‑related operating costs (e.g., token consumption) as a constraint, yet 60 % plan to increase AI investment over the next year.
- McKinsey warns that vendors must rethink the traditional “buy” model as enterprises become more deliberate about where to buy, build, or develop internal capabilities.
Overview of the McKinsey Findings
The McKinsey report, released on September 7 2026, highlights a pivotal shift in corporate technology strategy driven by the rise of agentic AI coding tools. According to the survey, “nearly one‑third (32 per cent) of respondents said their organisations had decided against purchasing at least one software product or feature because they could build the functionality internally using agentic coding tools.” This statistic underscores a growing confidence among enterprises that internal development, augmented by AI, can rival—or even surpass—off‑the‑shelf solutions in speed, customization, and cost‑effectiveness. The findings suggest that the long‑standing “buy versus build” debate is tilting decisively toward the latter, at least for certain classes of software.
Sector‑Specific Adoption Patterns
When broken down by industry, the inclination to build rather than buy is most pronounced in the technology and healthcare sectors. The report notes that these two fields lead the pack, “followed by professional services and energy and materials.” In technology firms, the proximity to software engineering talent and a culture of rapid experimentation make internal AI‑assisted development a natural fit. Healthcare organisations, meanwhile, are leveraging coding agents to create bespoke clinical‑data platforms and regulatory‑compliance tools that generic vendors often struggle to tailor. Professional services firms are using the technology to automate proprietary consulting methodologies, while energy and materials companies apply it to optimize supply‑chain simulations and predictive maintenance scripts.
Scaling Agentic AI in Large Enterprises
Large organisations—those with more than $1 billion in annual revenue—are accelerating their adoption of agentic AI at a notable pace. The survey reveals that “forty percent of respondents at organisations with more than $1 billion in annual revenue said they were scaling AI agents in at least one function, up from 27 per cent a year earlier.” This upward trajectory indicates that scaling is no longer a pilot‑phase experiment but a strategic initiative embedded in broader digital‑transformation roadmaps. By contrast, adoption among smaller organisations has remained “broadly unchanged at 22 per cent,” suggesting that resource constraints, limited AI expertise, or concerns about return on investment continue to hinder wider diffusion outside the enterprise segment.
AI High Performers Lead the Way
A distinct subset of respondents—identified as “AI high performers”—are outpacing their peers in both the depth and breadth of AI agent deployment. According to McKinsey, “they are twice as likely as other organisations to report scaling software coding agents and 2.7 times more likely to scale other agentic AI.” Furthermore, “nearly half of high performers said they had forgone software purchases in favour of building internally, compared with 31 per cent of other respondents.” These high performers typically exhibit a combination of strong data governance, mature MLOps practices, and a culture that rewards rapid experimentation. Their success illustrates that the benefits of agentic AI are not merely technological but also organisational, hinging on the ability to integrate AI workflows into core business processes.
Economic Considerations and Constraints
While enthusiasm for AI‑driven development is high, cost considerations are beginning to surface as a practical limiting factor. The report states that “around one in five organisations said AI‑related operating costs, including token costs, had constrained their use of the technology.” Token consumption—particularly when leveraging large‑language‑model APIs for code generation—can accumulate quickly, especially for organisations running continuous integration pipelines that invoke the model thousands of times per day. Nevertheless, investment appetite remains robust: “60 per cent of respondents expecting their organisations to increase AI investment over the next year.” This dichotomy suggests that firms are willing to absorb higher operating expenses in exchange for the strategic advantages of custom software, reduced vendor lock‑in, and accelerated time‑to‑market.
Implications for Technology Vendors
The growing propensity to build internally poses a fundamental challenge to traditional software vendors, whose business models have historically relied on licensing or subscription revenues. McKinsey advises that vendors must “rethink the traditional software ‘buy’ model, particularly as coding agents become more capable.” To remain relevant, vendors may need to pivot toward offering AI‑ready platforms, low‑code environments, or consultancy services that help customers harness agentic AI effectively. Additionally, providing transparent pricing models for token usage and offering cost‑optimisation tools could mitigate one of the key concerns highlighted by surveyed organisations. In essence, vendors must evolve from pure product sellers to enablers of internal AI‑driven development.
Future Outlook and Strategic Advice
Looking ahead, the trajectory indicated by the McKinsey data points to a hybrid landscape where buying, building, and internal capability development coexist, with decisions increasingly driven by specific use‑case economics rather than blanket preferences. Organisations are urged to adopt a deliberate framework: evaluate whether a given function benefits more from the speed and customisation of AI‑assisted internal build, or from the reliability, support, and continual updates offered by established vendors. Investment in AI literacy, robust MLOps pipelines, and cost‑monitoring mechanisms will be critical to scaling successfully while keeping operating expenses in check. As agentic AI matures, the line between “buy” and “build” will likely blur further, compelling both enterprises and vendors to innovate continuously in order to capture the next wave of value.
https://cio.economictimes.indiatimes.com/news/artificial-intelligence/ai-coding-agents-shift-corporate-tech-spending-companies-build-software-internally-over-buying/133859948