As Python draws strong interest as a programming language for enterprise AI development, some argue Java may be a better choice.

Bruno Borges (브루노 보르헤스), a principal product and community manager in charge of Java at Microsoft, is among those who take that view.

According to a report by The New Stack, he said, "Compared with other language runtimes such as Python and Node.js, the JVM is far more cost-efficient" in terms of performance and efficiency.

Lower runtime costs in AI development typically mean more budget for tokens and API calls. The argument is that an efficient Java runtime can help save computing resources when operating hundreds to thousands of AI agents in an enterprise environment.

Frameworks that support AI development in the Java ecosystem are also emerging. Notable examples include LangChain4j, which simplifies LLM integration, Spring AI and embabel, which adds agent flows.

Julien Dubois (줄리앙 뒤부아), head of Microsoft’s developer relations team for Java, said, "AI needs tools, databases and MCP servers, and Java has strengths in integrating third-party solutions."

As AI writes code, the criteria for choosing a programming language are changing. Borges stressed that when AI agents write most code, the key criterion is readability, not the shortest code. He believes Java’s explicit and detailed syntax makes it easier for developers to understand and review code suggested by AI coding assistants.

The New Stack also reported that AI coding tools such as GitHub Copilot, Claude Code and Cursor are seen as a strength because they can effectively write code for popular Java frameworks such as Spring Boot and Hibernate.