Anthropic Engineer: AI Will Phase Out ‘Software Engineer’ Title by End of 2026

In a San Francisco lab that feels more like a Bell Labs revival than a typical tech startup, Anthropic engineer Evan Cherny made a prediction that landed like a thunderclap across the developer community. He hasn’t manually written a single line of code since November 2025. Every production commit he makes comes from Claude Code, Anthropic’s AI coding agent. And he’s convinced the transformation is only beginning.

“I think by the end of the year, everyone is going to be a product manager, and everyone codes,” Cherny told Fortune in a recent interview. “The title software engineer is going to start to go away. It’s just going to be replaced by ‘builder,’ and it’s going to be painful for a lot of people.”

“I haven’t manually edited a single line of code since November 2025. This is how I feel where I don’t have to do the tedious work anymore of coding.” — Evan Cherny, Claude Code Creator

The Tool Behind the Prediction

Cherny developed Claude Code while working in Anthropic’s experimental division, a research-oriented group modeled after the legendary Bell Labs. Released in 2025, the agentic system represents a fundamental shift in how software gets built. Unlike traditional coding assistants that suggest completions or answer questions, Claude Code autonomously executes tasks with minimal human intervention.

The system lives in the terminal, understands entire codebases, and can handle everything from building features to debugging, refactoring, and managing git workflows. Developers describe it as having a senior engineer who never sleeps, never gets tired, and can explain its reasoning when asked.

The numbers are striking. According to recent data, Claude Code now accounts for approximately 4% of all commits on GitHub. Anthropic’s annualized revenue has accelerated sharply since what industry observers call the “Claude Code Moment,” with the company’s valuation reaching $380 billion following a $30 billion Series G funding round in early 2026.

From Coding to Orchestrating

Cherny’s workflow illustrates the shift he believes will become universal. Rather than typing syntax, he describes problems in natural language, reviews Claude Code’s proposed solutions, and focuses on higher-level decisions about architecture and user needs.

“The fun part is figuring out what to build, and coming up with this,” he explained. “It’s talking to users. It’s thinking about these big systems. It’s thinking about the future. It’s collaborating with other people on the team, and that’s what I get to do more of now.”

“The fun part is figuring out what to build. It’s talking to users. It’s thinking about these big systems. That’s what I get to do more of now.” — Evan Cherny

This transition—from writing code to orchestrating AI agents—has profound implications for the profession. Cherny predicts that within a year or two, understanding core technical principles won’t matter as much as the ability to articulate problems clearly and evaluate solutions critically.

The Cowork Expansion

Anthropic isn’t limiting this transformation to experienced developers. In early February 2026, the company launched Claude Cowork, a more accessible AI agent designed for semi-technical users and non-coders. The tool handles daily management and organizational tasks without requiring programming knowledge.

Cowork represents Anthropic’s bet that AI agents will expand into “pretty much any kind of work that you can do on a computer.” The company has also released nine free Claude Skills tutorials covering Excel workflows, Chrome browsing, file editing, and project management—enabling beginners to build functional agents in under an hour.

The competitive landscape is intensifying. OpenAI’s Codex, powered by GPT-5.2-Codex, offers cloud-based agents with IDE extensions and open-source CLI tools. Google DeepMind continues advancing its Gemini models for coding tasks. But Anthropic’s terminal-first approach and focus on deep codebase understanding have carved out a distinctive position, particularly among developers who value thoroughness over speed.

The Industry Reckoning

Cherny’s prediction about job titles isn’t merely speculative—it’s based on patterns already visible in how teams are restructuring. Companies are increasingly hiring for “AI product managers” and “AI operations specialists” rather than traditional software engineers. The emphasis is shifting from technical implementation to problem definition and solution validation.

The transition won’t be painless. Cherny acknowledges that many developers will struggle with the identity shift. Years of investment in learning specific languages, frameworks, and tools may feel devalued when an AI can generate equivalent code in seconds.

But the alternative—resisting the change—may be riskier. As Claude Code and similar tools become standard, the developers who thrive will likely be those who learn to partner effectively with AI agents, focusing on the creative and strategic work that remains stubbornly human.

“In a year or two, it’s not going to matter. Everyone codes.” — Evan Cherny on the future of technical skills

For now, Cherny continues his code-free development workflow, treating Claude Code as both tool and teammate. Whether his prediction about job titles proves accurate by December 2026 remains to be seen. But the direction of travel seems clear: the era of humans as primary code authors is ending, and the era of humans as AI orchestrators is beginning.

This article was reported by the ArtificialDaily editorial team. For more information, visit Fortune and Anthropic.