AI-Native Software Engineering Creates the 'Agentic Engineer' Role

A new arXiv paper by Mamdouh Alenezi, 'From Determinism to Delegation: AI-Native Software Engineering and the Evolution of the Agentic Engineer,' argues that software engineering is undergoing its most significant transformation since high-level programming languages. The paper claims this is not just a tooling upgrade but a full paradigm shift, creating a new professional archetype: the Agentic Engineer, whose primary artifact is the agentic system (a supervised, multi-step AI workflow) rather than a traditional program. Alenezi characterizes the transition through three key changes: (i) the unit of work shifts from writing individual functions to designing and supervising agent workflows that use LLMs, tools, and memory; (ii) correctness shifts from binary assertions (pass/fail) to statistical evaluation under uncertainty; and (iii) accountability shifts from code authorship to outcome ownership—the engineer owns what the agent produces, not just the instructions.

Empirical evidence is mixed: some studies report productivity gains from AI agents, but others show experienced developers experiencing slowdowns, suggesting that disciplined oversight—not full automation—is the critical new competency. The paper also examines risks like indirect prompt injection and behavioral drift, and places human-AI collaboration within socio-technical frameworks. Alenezi concludes that the future is symbiosis rather than substitution: agentic engineering builds upon and depends on classical software engineering principles, meaning engineers must learn to delegate effectively while maintaining rigorous supervision.