DEV introduced AI Disclosure on August 26, 2026, giving authors three structured labels for how AI tools were used in posts: Hand Written (No AI), AI-Assisted (Some AI), and Fully Autonomous. Ben Halpern said the feature is meant to add transparency and nuance while DEV revises its AI guidelines and adds tools that give more weight to human curation by community leaders.

DEV said human connection, authentic discussion, and peer-to-peer knowledge sharing remain central to the community, even as writers use AI tools to draft posts, generate code samples, make major copyedits, translate text, or work with agents while coding. The platform also allows some autonomous posting by creatively developed bots, but only when it is not deceptive and readers receive enough context to judge what they are reading.

The disclosure tiers separate fully human work from AI-assisted writing and content generated primarily or entirely by automated agents or LLMs. DEV said an author who uses an LLM while debugging, generates boilerplate, or asks an agent to outline notes should mark the post as AI-Assisted and remains accountable for accuracy and quality. It gave examples of transparent autonomous bots, including systems that summarize CVE security alerts, track open-source release logs, or compile benchmark digests.

DEV said disclosure does not weaken its community standards. Low-effort, undisclosed AI-generated content or synthetic output presented as firsthand human experience violates guidelines and may lead to account suspension. The rollout is framed as a foundation for clearer expectations and fair standards, not an overnight change in moderation enforcement.

DEV is also building Feed Preferences so readers can reduce or filter AI-assisted or autonomous posts from their home feed. An “AI Content in Feed” option is already available in settings, and DEV said its exact effect will evolve over time. The implementation can be inspected because DEV’s codebase is open source through Forem.

DEV introduced AI Disclosure on August 26, 2026, giving authors three structured labels for how AI tools were used in posts: Hand Written (No AI), AI-Assisted (Some AI), and Fully Autonomous. Ben Halpern said the feature is meant to add transparency and nuance while DEV revises its AI guidelines and adds tools that give more weight to human curation by community leaders.

DEV said human connection, authentic discussion, and peer-to-peer knowledge sharing remain central to the community, even as writers use AI tools to draft posts, generate code samples, make major copyedits, translate text, or work with agents while coding. The platform also allows some autonomous posting by creatively developed bots, but only when it is not deceptive and readers receive enough context to judge what they are reading.

The disclosure tiers separate fully human work from AI-assisted writing and content generated primarily or entirely by automated agents or LLMs. DEV said an author who uses an LLM while debugging, generates boilerplate, or asks an agent to outline notes should mark the post as AI-Assisted and remains accountable for accuracy and quality. It gave examples of transparent autonomous bots, including systems that summarize CVE security alerts, track open-source release logs, or compile benchmark digests.

DEV said disclosure does not weaken its community standards. Low-effort, undisclosed AI-generated content or synthetic output presented as firsthand human experience violates guidelines and may lead to account suspension. The rollout is framed as a foundation for clearer expectations and fair standards, not an overnight change in moderation enforcement.

DEV is also building Feed Preferences so readers can reduce or filter AI-assisted or autonomous posts from their home feed. An “AI Content in Feed” option is already available in settings, and DEV said its exact effect will evolve over time. The implementation can be inspected because DEV’s codebase is open source through Forem.