[Policy Proposal] Revise Generative AI Policy to Allow AI-Assisted Contributions with Disclosure

[ylevental]

Summary

I would like to propose a revision to the current Generative AI policy to allow AI-assisted contributions with proper disclosure.

First, I apologize for not being aware of this policy before my previous contributions. However, I believe this presents an opportunity to reconsider the policy in light of recent developments.

Rationale for Policy Change

1. Demonstrated Quality Through Peer Review

AI-generated content has become significantly more accurate and reliable. As evidence, I have submitted ten peer-reviewed articles that were created with AI assistance, all of which were merged after passing your content team's review process:

View merged contributions

These contributions met your standards for technical accuracy, formatting, and quality—demonstrating that AI-assisted content can satisfy the same rigorous requirements as traditionally written content.

2. Authorship Credit with Transparency

The current policy cites plagiarism concerns regarding authorship credit. I propose that contributors be allowed to claim authorship when they:

• Disclose the use of AI tools in their contribution
• Review and verify the accuracy of AI-generated content
• Take responsibility for the final submission

This approach maintains integrity while acknowledging the contributor's role in curating, verifying, and submitting quality content.

3. Evolving Educational Paradigms

While Docs serves as an educational space for learning technical writing, the landscape of technical documentation is evolving. Modern technical writers increasingly use AI as a productivity tool. Teaching contributors to effectively leverage AI—while maintaining quality standards—may be a more relevant skill than prohibiting its use entirely.

The goal of creating accurate, helpful documentation can coexist with the use of modern tools, just as we don't require contributors to avoid IDEs with autocomplete or spell-checkers.

4. Productivity and Accessibility Benefits

Allowing AI-assisted contributions could:

• Lower barriers for non-native English speakers
• Enable faster creation of comprehensive documentation
• Allow contributors to focus on technical accuracy and verification
• Increase the volume of quality contributions to the project

Proposed Policy Changes

1. Allow AI-assisted contributions with mandatory disclosure
2. Require contributors to verify all AI-generated content for accuracy
3. Maintain existing quality standards for all submissions, regardless of creation method
4. Add a disclosure field in PR templates for AI usage

Code of Conduct

By submitting this issue, I agree to follow Codecademy Docs' Code of Conduct.

---

I welcome discussion on this proposal and am happy to provide additional context or address any concerns.

[ylevental]

@mamtawardhani Remember, technology is evolving all the time! This would be a great opportunity to embrace the current uptick in AI accuracy and usage 😄

To everyone else, please be civil in discussing this important issue. It's required by GitHub.

[StuartMosquera]

If a contributor needs to use AI to write a short article about a technology they're supposed to know, the problem is obvious. I'm not saying that using AI is bad; actually, the Contribution Guide says that tools like Grammarly are allowed for grammar checks.

The content is supposed to be written by a person; what’s the point of a contribution if it’s entirely generated by AI?

[cryptiklemur]

Agreed. Using AI to write educational content is entirely disingenuous, and in my honest opinion, is just an attempt by the author to gain reputation/clout.

[ylevental]

Both of you make good points, but what does "exclusively" mean? Where do we draw the line, and what if someone just uses AI and doesn't say anything?

The contribution could still help other people, AI-generated or not.

Also, where do we draw the line with technology in general? One could argue that computers themselves make things too simplistic, and we should type all documentation with a typewriter.

[cryptiklemur]

Its pretty easy to draw the line between asking Claude to write you an entire post, and using tools like Grammarly to improve grammer and syntax. The later doesn't change substance, just structure.

This shouldn't be some meta-ethical debate.

[ylevental]

These are small purely technical articles, not the Great American Novel 😉

They would be easy to create through copy-pasting and slight rewording... which people have been doing for a long time anyway

Of course, writing cultural works using AI would create serious ethical issues

[cryptiklemur]

They would be easy to create through copy-pasting and slight rewording... which people have been doing for a long time anyway

This is pure conjecture, on both points.

Of course, writing cultural works using AI would create serious ethical issues

So does this, just with different arguments.

[StuartMosquera]

These are small purely technical articles, not the Great American Novel 😉

You've made my point here; if they're small, what's the point of using AI to write their entire content? 😆

[ylevental]

These are small purely technical articles, not the Great American Novel 😉

You've made my point here; if they're small, what's the point of using AI to write their entire content? 😆

It could be good for writing articles en masse. The only noticeable difference would be speed.

[cryptiklemur]

How could it be good...? Other than to raise one's "clout", it serves no purpose.

[StuartMosquera]

@mamtawardhani Is the use of AI already allowed or not? I'm asking because in PR #7970 you didn't answer ylevental's question.

[ylevental]

Hello all,

It looks like the CONTRIBUTING.md guidelines were updated just a few hours ago to address this!

Can I use AI tools when contributing?
"Yes! You can use AI tools to help create our Docs entries... AI can assist you with drafting, rewriting, or structuring content. However, you are responsible for ensuring your work meets Codecademy’s quality and style standards."
Link to updated guidelines

This perfectly clarifies the stance: AI is allowed as a productivity tool, provided the contributor takes full accountability for the accuracy and quality of the final output.

Thank you to the maintainers for this update. Since this resolves the proposal, I will go ahead and close this issue. Future discussion should be raised in a separate issue.

[github-actions]

🌸 Thanks for closing this Issue!