Oracle bans AI-generated code in OpenJDK over legal and safety fears
Even corporate giants hyping machine learning behind closed doors are drawing a hard line when AI starts scribbling directly into foundational software.
The maintainers behind OpenJDK officially updated their legal guidelines to block any incoming code, documentation, or graphics created by generative models. Developers are strictly forbidden from submitting AI-crafted commits to repositories, wiki pages, or mailing lists, leaving automated tools relegated strictly to code review and debugging duties.
The sudden crackdown stems from a toxic cocktail of copyright ambiguity and security nightmare scenarios. Because large language models suck up vast amounts of public repositories during training, nobody actually knows who owns the resulting output, raising immediate red flags over potential licensing lawsuits down the line. To make matters worse, machine-generated patches are increasingly plagued by hallucinations and silent security vulnerabilities that human reviewers are tired of babysitting.
This anti-bot backlash isn't an isolated incident in the open-source ecosystem. Greg Kroah-Hartman, a key maintainer for the Linux kernel, previously banned AI-generated patches in the staging driver section. That specific area exists as a messy sandbox for human rookies to practice fixing real bugs, and flooding it with automated pull requests completely defeated the purpose of training actual software engineers.
When tech giants who market machine intelligence as the ultimate replacement for human labor suddenly lock their own front doors against it, the hype machine hits a wall of cold reality. Software architecture built on legal liability and unvetted code hallucinations is apparently a bridge too far even for those selling the dream.
Source: The Register
Comments
This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.