Startups race ahead in adopting AI for coding while enterprises play catch-up

"These adoption patterns mirror past technology shifts, where startups use new tools for competitive advantage while established organizations move more cautiously and often have detailed security checks in place before adopting new tools company-wide," it added.

While startups stood out as the fastest adopters, other major user groups include students, academics, and individual hobbyists.

"If AI agents provide significant productivity gains, the gap between early and late adopters could translate into substantial competitive advantages," the report said.

What developers are building

AI-assisted coding is being used extensively for building websites and apps, according to the report.

Languages like JavaScript and TypeScript accounted for 31 percent of all programming language mentions across the platforms, while HTML and CSS added another 28 percent.

User interface and user experience (UI/UX) development was one of the most common task categories, accounting for 12 percent of all interactions, followed by web and mobile app development at 8 percent.

“Developers commonly use Claude for building user interfaces and interactive elements,” the report noted, adding that this trend may signal early disruption for jobs focused on building simple apps and front-end code.

The trend coincides with the rise of “vibe coding”, where developers describe outcomes in natural language and rely on AI to generate the actual code.

While front-end development dominated, backend and data tasks weren't missing. Python made up 14 percent of usage, covering both backend and data science tasks, while SQL accounted for 6 percent.

Automation is taking over

The shift is not just in what developers build with AI, but how.

According to the report, 79 percent of conversations on Claude Code were identified as “automation”, where AI directly performs tasks rather than “augmentation,” where AI collaborates with and enhances human capabilities (21 percent). In contrast, only 49 percent of Claude.ai conversations were classified as automation.

"This might imply that as AI agents become more commonplace, and as more agentic AI products are built, we should expect more automation of tasks," the report said.

“Feedback Loop” patterns, where Claude completes tasks autonomously but with the help of human validation, were nearly twice as common on Claude Code (35.8 percent of interactions) as Claude.ai (21.3 percent), the report said.

“Directive” conversations, where Claude completed a task with minimal user interaction, were also higher on Claude Code (43.8 percent, versus 27.5 percent on Claude.ai), it added.

What next?

With agents like Claude Code increasingly capable of writing, testing, and improving software autonomously, the developer’s role may shift from hands-on coding to quality control.

"As AI systems become capable of building larger-scale pieces of software, will developers shift to mostly managing and guiding these systems, rather than writing code themselves? Which software development roles will change the most, and which might disappear entirely?" the report questioned.

As AI gains traction in more industries, coding could serve as an indicator for how other white-collar jobs might evolve.