A policy proposal on our approach to deepfake tools and responsible AI

GitHub proposes new policy on deepfake and manipulated media tools

GitHub has opened a public comment period on a proposed addition to its Acceptable Use Policies that would restrict projects designed to create synthetic or manipulated media for harm. The change, published as a pull request in the company's site-policy repository, would specifically target tools used to produce non-consensual intimate imagery (NCII) or content that violates the platform's existing misinformation and disinformation policies.

Synthetic and Manipulated Media Tools
GitHub does not allow any projects that are designed for, encourage, promote, support, or suggest in any way the use of synthetic or manipulated media for the creation of non-consensual intimate imagery or any content that would constitute misinformation or disinformation under this policy.

Policy context: AI abuse vectors are multiplying

The proposal arrives as AI-generated content becomes harder to distinguish from reality. GitHub points to the approaching global election cycle and growing concern over deepfake videos, images, and robocalls designed to mislead voters. At the same time, high-profile NCII cases involving synthetic media have driven a wave of state and federal legislative proposals in the United States.

Because GitHub hosts source code that could be used to build these tools, the company says it has a responsibility to address the abuse potential while still protecting legitimate research and creative work. The proposed policy is framed as an update to keep platform rules aligned with the current state of the technology.

Balancing open source values against misuse

The core tension for GitHub is that its mission depends on keeping code publicly available. Removing code from the platform removes educational value and limits the ability of security researchers and others to study how these systems work. GitHub previously faced a similar question when updating its malware and exploit policies, where it chose to disallow actively harmful projects while permitting dual-use content shared for legitimate security research.

The proposed synthetic media policy takes a similar dual-use stance. Projects that contribute to public knowledge or enable creative expression remain welcome, provided they are not designed to produce NCII or disinformation. The policy distinguishes between tools whose primary purpose is harmful and those with legitimate general use, with enforcement responses intended to be proportional to the severity of the violation.

Public comments open until May 20

The comment window for the proposed policy addition runs for 30 days, closing May 20. Interested parties can review the full text and submit feedback through the pull request in the github/site-policy repository, where GitHub routinely collaborates with the community on policy revisions.