Robotic process automation keeps getting declared dead, and this week it happened again. In a Forbes Technology Council piece published August 26, a senior application development executive at First Citizens Bank pushed back hard against the idea that generative AI and autonomous agents have made RPA obsolete. As the author put it, with the rapid rise of Generative AI and autonomous agents, a common narrative has emerged that RPA is dead, but the author doesn't believe that's true. The piece is notable less for breaking news and more for how it crystallizes a debate that has been simmering across the enterprise software world all year: is RPA a legacy technology being phased out, or is it the durable foundation that agentic AI actually needs to work?

The numbers suggest the latter, at least for now. A market research report circulated in early August pegged the global RPA market at roughly $6.1 billion in 2025, projecting growth to $7.7 billion in 2026 and all the way to $62 billion by 2035 — a 26.1% compound annual growth rate. Cloud licensing and hyperscaler marketplace access through AWS, Azure, and Google Cloud Marketplace have removed the capital-expenditure barrier that historically limited RPA adoption to large enterprises. More strikingly, the report frames regulation as a growth engine rather than a headwind: the EU's Digital Operational Resilience Act, which mandates immutable audit trails for financial institutions, is one of two catalysts driving the market's trajectory. In other words, the same compliance pressures pushing banks and hospitals toward auditability are keeping deterministic, rule-based bots relevant even as flashier agentic tools grab headlines.

That tension — auditability versus autonomy — is exactly what vendors have spent 2026 trying to resolve, not by abandoning RPA but by wrapping it in orchestration layers. SS&C Blue Prism, whose founders helped coin the term "RPA" over fifteen years ago, has been positioning its WorkHQ platform around exactly this idea: treating orchestration — how businesses deploy and manage AI agents and traditional automation together, with governance and control around them — as one of the most pressing operational problems automation must solve. The pitch isn't "replace your bots," it's "give your bots a smarter, more accountable supervisor."

Taken together, these threads point to a maturing consensus rather than a clean winner. Rule-based bots remain the cheapest, most auditable way to move data through legacy systems that lack APIs, especially in regulated industries. Agentic AI, meanwhile, is best suited for the messy, unstructured 80% of enterprise work — emails, PDFs, judgment calls — that traditional RPA was never built to handle. The real story isn't extinction; it's stratification, with RPA increasingly treated as the reliable execution layer beneath a new generation of reasoning agents.

Sources

AI AgentsAgentic AIDaily RPA UpdateEnterprise AutomationIntelligent AutomationRPA