Draft:AI Pair Coding

Submission declined on 4 December 2025 by Pythoncoder (talk).

Your draft shows signs of having been generated by a large language model, such as ChatGPT. Wikipedia guidelines prohibit the use of LLMs to write articles from scratch. In addition, LLM-generated articles usually have multiple quality issues, to include:

Promotional tone, editorializing and other words to watch
Vague, generic, and speculative statements extrapolated from similar subjects
Essay-like writing
Hallucinations (plausible-sounding, but false information) and non-existent references
Close paraphrasing

Please address these issues. The best way is usually to read reliable sources and summarize them, instead of using a large language model. See our help page on large language models.

If you would like to continue working on the submission, click on the "Edit" tab at the top of the window.
If you have not resolved the issues listed above, your draft will be declined again and potentially deleted.
If you need extra help, please ask us a question at the AfC Help Desk or get live help from experienced editors.
Please do not remove reviewer comments or this notice until the submission is accepted.

Where to get help

If you need help editing or submitting your draft, please ask us a question at the AfC Help Desk or get live help from experienced editors. These venues are only for help with editing and the submission process, not to get reviews.
If you need feedback on your draft, or if the review is taking a lot of time, you can try asking for help on the talk page of a relevant WikiProject. Some WikiProjects are more active than others so a speedy reply is not guaranteed.

How to improve a draft

Wikipedia:Contributing to Wikipedia – a basic overview on how to edit Wikipedia.
Help:Wikitext – how to use the markup
Help:Referencing for beginners – how to include references
Wikipedia:Article development – how to develop your article
Wikipedia:Writing better articles – how to improve your article
Wikipedia:Verifiability – make sure your article includes reliable third-party sources

You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed article.

Improving your odds of a speedy review

To improve your odds of a faster review, tag your draft with relevant WikiProject tags using the button below. This will let reviewers know a new draft has been submitted in their area of interest. For instance, if you wrote about a female astronomer, you would want to add the Biography, Astronomy, and Women scientists tags.

Add tags to your draft

Editor resources

Easy tools: Citation bot (help) | Advanced: Fix bare URLs

Declined by Pythoncoder 15 days ago. Last edited by Keith D 15 days ago. Reviewer: Inform author.

Resubmit

Please note that if the issues are not fixed, the draft will be declined again.

AI Pair Coding (APC) is a software development practice where developers actively review, test, understand, and take full ownership of code generated by agentic AI systems such as GitHub Copilot, Claude Code, or similar AI coding assistants.^[1] The term was coined by Andrew Penrose from IBM in 2025 to distinguish this rigorous, review-focused approach from the broader concept of AI pair programming.^[2] The practice emphasizes human oversight and professional accountability throughout the development process.

Definition and practice

AI Pair Coding treats artificial intelligence as a collaborative partner in software development, where the human developer maintains responsibility for code quality, architectural decisions, security review, and maintainability.^[3] The approach requires developers to understand the generated code before integration, run tests to verify correctness, and ensure the code meets professional standards.

Key characteristics of AI Pair Coding include:^[4]

Active code review of all AI-generated suggestions
Testing and validation before deployment
Understanding code architecture and logic
Maintaining professional development standards
Taking full ownership of the final codebase

Distinction from AI pair programming

While "AI pair programming" is a general term used by tool vendors to describe any AI-assisted coding,^[5] AI Pair Coding specifically emphasizes the developer's active engagement with rigorous review practices. Penrose introduced the term to create a distinction between professional, accountability-driven development practices and more casual AI code generation approaches.

Distinction from vibe coding

AI Pair Coding is explicitly contrasted with vibe coding, a term coined by OpenAI co-founder Andrej Karpathy in February 2025.^[6] While vibe coding describes accepting AI-generated code without review or understanding,^[7] AI Pair Coding emphasizes thorough review and comprehension.

Developer Simon Willison, co-creator of the Django framework, distinguished the practices: "If an LLM wrote the code for you, and you then reviewed it, tested it thoroughly and made sure you could explain how it works to someone else that's not vibe coding, it's software development."^[8]

Tools and adoption

Several agentic AI systems support AI Pair Coding practices:

GitHub Copilot - Launched in 2021, described by GitHub as "an AI pair programmer,"^[9] with over 20 million users as of 2025
Claude Code - A command-line agentic coding tool by Anthropic that integrates with developer workflows^[10]^[11]
IBM Project Bob - An AI coding system combining multiple LLMs with agentic review capabilities^[12]

Industry analyst Gartner projected that 75% of enterprise software engineers would use AI code assistants by 2028.^[13]

Security and quality considerations

Research has identified security concerns with AI-generated code that AI Pair Coding practices aim to address. A 2024 study found that 48% of code snippets from LLMs contained exploitable bugs,^[14] while security firm Endor Labs reported over 40% of AI-generated code contains security flaws.^[15]

Industry guidance emphasizes human review as essential. GitHub's responsible AI documentation states: "The code should still be run and tested locally. And of course, code reviews should not be skipped!"^[16]

Productivity research

A 2023 study by Microsoft Research and GitHub found developers using AI pair programming tools completed tasks 55.8% faster than control groups.^[17] However, research also indicates that proper implementation of AI Pair Coding practices—including code review and testing—is essential to realizing productivity gains while maintaining code quality.

References

^ "Responsible AI pair programming with GitHub Copilot". GitHub. Retrieved 2023. {{cite web}}: Check date values in: |access-date= (help)
^ Andrew Penrose (2025). "Introducing AI Pair Coding (APC): A Term We Actually Need". LinkedIn.
^ "Developers with AI assistants need to follow the pair programming model". Stack Overflow. 2024-04-03.
^ "AI code review implementation and best practices". Graphite.
^ "Pair Programming with Git Copilot". Cubet. Retrieved 2024. {{cite web}}: Check date values in: |access-date= (help)
^ "What is vibe coding, exactly?". MIT Technology Review. 2025-04-16.
^ Simon Willison (2025-03-19). "Not all AI-assisted programming is vibe coding (but vibe coding rocks)".
^ Simon Willison (2025-03-19). "Not all AI-assisted programming is vibe coding (but vibe coding rocks)".
^ "GitHub Copilot: Meet the new coding agent". GitHub.
^ "Claude Code". Anthropic.
^ "Claude Code GitHub Repository". GitHub.
^ "IBM Project Bob". IBM.
^ "Gartner Says 75% of Enterprise Software Engineers Will Use AI Code Assistants by 2028". Gartner. 2024-04-11.
^ "Security risks of AI-generated code and how to manage them". TechTarget.
^ "The Most Common Security Vulnerabilities in AI-Generated Code". Endor Labs.
^ "Responsible AI pair programming with GitHub Copilot". GitHub.
^ "The Impact of AI on Developer Productivity: Evidence from GitHub Copilot". arXiv. February 2023.

[1] "Responsible AI pair programming with GitHub Copilot". GitHub. Retrieved 2023. {{cite web}}: Check date values in: |access-date= (help)

[2] Andrew Penrose (2025). "Introducing AI Pair Coding (APC): A Term We Actually Need". LinkedIn.

[3] "Developers with AI assistants need to follow the pair programming model". Stack Overflow. 2024-04-03.

[4] "AI code review implementation and best practices". Graphite.

[5] "Pair Programming with Git Copilot". Cubet. Retrieved 2024. {{cite web}}: Check date values in: |access-date= (help)

[6] "What is vibe coding, exactly?". MIT Technology Review. 2025-04-16.

[7] Simon Willison (2025-03-19). "Not all AI-assisted programming is vibe coding (but vibe coding rocks)".

[8] Simon Willison (2025-03-19). "Not all AI-assisted programming is vibe coding (but vibe coding rocks)".

[9] "GitHub Copilot: Meet the new coding agent". GitHub.

[10] "Claude Code". Anthropic.

[11] "Claude Code GitHub Repository". GitHub.

[12] "IBM Project Bob". IBM.

[13] "Gartner Says 75% of Enterprise Software Engineers Will Use AI Code Assistants by 2028". Gartner. 2024-04-11.

[14] "Security risks of AI-generated code and how to manage them". TechTarget.

[15] "The Most Common Security Vulnerabilities in AI-Generated Code". Endor Labs.

[16] "Responsible AI pair programming with GitHub Copilot". GitHub.

[17] "The Impact of AI on Developer Productivity: Evidence from GitHub Copilot". arXiv. February 2023.

[1]