|
| 1 | +title: Avoiding generative AI development tools |
| 2 | +--- |
| 3 | +categories: |
| 4 | +community |
| 5 | +open-source |
| 6 | +--- |
| 7 | +author: TimidRobot |
| 8 | +--- |
| 9 | +series: |
| 10 | +--- |
| 11 | +pub_date: 2025-12-01 |
| 12 | +--- |
| 13 | +body: |
| 14 | + |
| 15 | +At this time, the Creative Commons (CC) **Technology team** has concluded that |
| 16 | +generative AI development tooling does not pass a cost/benefit analysis *for |
| 17 | +contributing to our open source projects*. Until further notice, we will not |
| 18 | +accept CC Open Source submissions that include code or content generated with |
| 19 | +AI. |
| 20 | + |
| 21 | + |
| 22 | +## Overhyped |
| 23 | + |
| 24 | +Generative AI productivity claims are often championed (and paid for) by those |
| 25 | +who reap a material benefit from adoption of AI tooling. Adoption rates are |
| 26 | +inflated by a lack of choice when AI tools are included by default and often |
| 27 | +cannot be disabled (EditorDavid, 2025). We’ve observed that arguments for their |
| 28 | +adoption too often appeal to a fear of missing out or potential future |
| 29 | +benefits. |
| 30 | + |
| 31 | + |
| 32 | +## Inefficiency and slop |
| 33 | + |
| 34 | +Our current evaluation of generative AI development tooling is that it does not |
| 35 | +offer a benefit to our open source projects nor to the user of the AI tooling. |
| 36 | +Indeed, "when developers use AI tools, they take 19% longer than without—AI |
| 37 | +makes them slower." (METR, 2025). A recent report from MIT NANDA indicates |
| 38 | +"that 95% of organizations are getting zero return [from generative AI]." |
| 39 | +(Challapally et al., 2025). AI tooling enables the creation of spurious work |
| 40 | +for reviewers. For example, see the following comment on an AI slop |
| 41 | +submission: |
| 42 | + |
| 43 | +> This code does not call curl. This is not a "POC" of anything than [sic] |
| 44 | +> suggesting you did this with an AI and that you do not understand what you're |
| 45 | +> doing here. |
| 46 | +> <br>Please identify the exact line of curl source code that has the problem |
| 47 | +> you say exists. |
| 48 | +> <br>(badger, 2025) |
| 49 | +
|
| 50 | + |
| 51 | +## Counterproductive |
| 52 | + |
| 53 | +Especially for our contributors, when we participate in work programs like |
| 54 | +Google Summer of Code (GSoC) or Outreachy, AI tools train the wrong skills. We |
| 55 | +want contributors who are learning the technologies that underpin our open |
| 56 | +source projects. |
| 57 | + |
| 58 | +> I listened to an industry representative say that he wants students to have |
| 59 | +> “design thinking and the ability to ask questions” without recognizing that |
| 60 | +> the student use of AI tools he advocates will stunt their critical thinking |
| 61 | +> skills, making it impossible for them to ask questions and do interesting |
| 62 | +> “design thinking.”. |
| 63 | +> <br>(Valeries, 2025) |
| 64 | +
|
| 65 | +## Responsibility and solidarity |
| 66 | + |
| 67 | +In addition to concerns with how AI development tools might negatively impact |
| 68 | +our day-to-day, the Technology team has concerns about how generative AI |
| 69 | +technology is created and implemented/deployed. Even if generative AI |
| 70 | +development tools were actually helpful, the Technology team would be reluctant |
| 71 | +to use tools that were developed in a way that harmed others, that are funded |
| 72 | +by deployments/implementations that harm others, and that offer short-term |
| 73 | +benefits at an existential cost. |
| 74 | + |
| 75 | +## References |
| 76 | + |
| 77 | +- badger. (2025, September 16). Comment on Stack Buffer Overflow in cURL Cookie |
| 78 | + Parsing Leads to RCE. curl. |
| 79 | + [https://hackerone.com/reports/3340109#activity-36894924](https://hackerone.com/reports/3340109#activity-36894924) |
| 80 | +- Challapally, A., Pease, C., Raskar, R., Chari, P., (2025, July). The GenAI |
| 81 | + Divide: STATE OF AI IN BUSINESS 2025. MIT NANDA. |
| 82 | + [https://www.artificialintelligence-news.com/wp-content/uploads/2025/08/ai_report_2025.pdf](https://www.artificialintelligence-news.com/wp-content/uploads/2025/08/ai_report_2025.pdf) |
| 83 | +- EditorDavid. (2025, May 31). GitHub Users Angry at the Prospect of AI-Written |
| 84 | + Issues From Copilot. Slashdot. |
| 85 | + [https://developers.slashdot.org/story/25/06/01/0049240/github-users-angry-at-the-prospect-of-ai-written-issues-from-copilot](https://developers.slashdot.org/story/25/06/01/0049240/github-users-angry-at-the-prospect-of-ai-written-issues-from-copilot) |
| 86 | +- METR. (2025, July 10). Measuring the Impact of Early-2025 AI on Experienced |
| 87 | + Open-Source Developer Productivity. METR. |
| 88 | + [https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/) |
| 89 | +- Valeries, Barr. (2025, August 29). Feeling Cranky About AI and CS Education. |
| 90 | + Communications of the ACM. |
| 91 | + [https://cacm.acm.org/blogcacm/feeling-cranky-about-ai-and-cs-education/](https://cacm.acm.org/blogcacm/feeling-cranky-about-ai-and-cs-education/) |
0 commit comments