Two related, Oracle-backed projects published opposing policies on open-source contributions created with generative AI: The OpenJDK Governing Board approved an interim policy prohibiting such contributions, while the Coding Assistants policy from GraalVM permits them. Both projects require contributors to sign the same Oracle Contributor Agreement (OCA) for intellectual property.
In early April 2026, OpenJDK published its policy that broadly bans generative AI content:
Contributions must not include content generated, in part or in full, by large language models, diffusion models, or similar deep-learning systems. Content, in this context, includes but is not limited to source code, text, and images in OpenJDK Git repositories, GitHub pull requests, e-mail messages, wiki pages, and JBS issues.
This policy provides three reasons. First, reviewer burden: floods of plausible-looking but incorrect or hard-to-maintain code drain limited reviewer time. Second, safety and security: the JDK underpins mission-critical systems, and that demands a high bar. Third, intellectual property: the OCA requires contributors to own the IP rights they grant to Oracle without restriction, but whether a person owns the IP rights in AI-generated output is "the subject of active litigation."
The policy carves out private use. Contributors may still use generative AI to understand, debug, and review OpenJDK code and for research — they just may not contribute AI-generated content. The policy FAQ is explicit: editing ten of 100 AI-generated lines is still not allowed because the contribution is still partly AI-generated. The policy also allows tools such as "spell-checking, grammar-checking, auto-completion, and refactoring features in my editor or IDE" if "they are not based on large language models or similar deep-learning systems."
OpenJDK contributors must soon check a checkbox in Skara, the automated pull request review system, to confirm their contributions comply with the generative AI policy. OpenJDK acknowledges that "in general, reliably distinguishing human-generated content from AI-generated content is impossible", but encourages reviewers to watch out for tell-tale signs of AI-generated content nevertheless.
In mid-April 2026, GraalVM, an Oracle Labs project not governed by the OpenJDK Governing Board, clarified their AI-assisted contribution policy and contributor guidance that allows generative AI content:
GraalVM contributors may use AI coding assistants and similar tools when preparing contributions. [...] For purposes of this document, "coding assistants" includes AI-based tools that help draft, transform, explain, review, or summarize code, tests, documentation, or commit text. This policy applies to contributions and project interactions prepared with such tools, including pull requests and issues filed with the project
The project added a "Documentation Terminology and Style Guide" for AI coding assistants on June 3, 2026.
The GraalVM policy "is informed by the Linux kernel's AI CodingAssistants policy" but adapts it. For instance, the Linux policy states that "contributions should include an Assisted-by tag." GraalVM, on the other hand, has a weaker stance: It says that "explicit attribution to a specific model or tool is optional," but the "disclosure of AI assistance is encouraged when it helps reviewers understand how a change was produced."
Contributor accountability is the core of GraalVM's contributor-responsibility rules: The human submitter is responsible for the entire contribution, including any part assisted by AI. They must review and understand it, verify its correctness, and answer the reviewer's questions without deferring to a tool: "If a contributor cannot explain, defend, or maintain an AI-assisted change, the contribution may be rejected."
Maintainer review remains unchanged: "Use of AI assistance does not create any presumption that a change is correct, review-ready, or exempt from normal scrutiny." Maintainers may ask about provenance, design intent, licensing, testing, or contributor understanding when reviewing contributions.
Both projects require contributors to sign the same OCA and grant Oracle unrestricted IP rights. But while OpenJDK cites unclear IP exposure of AI-generated content as grounds for outright prohibition, GraalVM treats contributor accountability as sufficient grounds for permission.
Oracle is working on a full AI contribution policy for OpenJDK and will propose it in "due course." At this time, there are no publicly available announcements on the evolution of the GraalVM AI contribution policy.