{"id":492,"date":"2026-04-13T10:58:23","date_gmt":"2026-04-13T09:58:23","guid":{"rendered":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/?p=492"},"modified":"2026-04-13T11:15:20","modified_gmt":"2026-04-13T10:15:20","slug":"ai-assisted-software-development","status":"publish","type":"post","link":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/2026\/04\/13\/ai-assisted-software-development\/","title":{"rendered":"AI-assisted software development"},"content":{"rendered":"\n

AI-assisted software development<\/strong>\u00a0is the use of\u00a0artificial intelligence<\/a>\u00a0(AI) to augment\u00a0software development<\/a>. It uses\u00a0large language models<\/a>\u00a0(LLMs),\u00a0natural language processing<\/a>,\u00a0intelligent agents<\/a>\u00a0and other AI technologies to assist\u00a0software developers<\/a>. It helps in a range of tasks of the\u00a0software development life cycle<\/a>, from initial\u00a0code generation<\/a>\u00a0to subsequent\u00a0debugging<\/a>,\u00a0testing<\/a>\u00a0and\u00a0documentation<\/a>.[1]<\/a><\/sup>\u00a0Agentic coding<\/strong>\u00a0denotes the use of more\u00a0autonomous<\/a>\u00a0AI agents<\/a>\u00a0for software development<\/p>\n\n\n\n

Technologies<\/h5>\n\n\n\n
Code generation<\/h6>\n\n\n\n

LLMs that have been trained on source code<\/a> repositories<\/a> are able to generate functional code<\/a> from natural language prompts<\/a>. Such models have knowledge of programming syntax<\/a>, common design patterns<\/a> and best practices<\/a> in a variety of programming languages<\/a><\/p>\n\n\n\n

Intelligent code completion<\/h6>\n\n\n\n

AI agents using pre-trained and fine-tuned LLMs can predict and suggest code completions<\/a> based on context, going beyond simple keyword matching to infer the developer’s intent and picture the broader structure of the developing codebase<\/a>. An analysis has shown that such use of LLMs significantly enhances code completion performance across several programming languages and contexts, and the resulting capability of predicting relevant code snippets based on context and partial input boosts developer productivity substantially<\/p>\n\n\n\n

Testing, debugging, code review and analysis<\/h6>\n\n\n\n

AI is used to automatically generate test cases<\/a>, identify potential bugs<\/a>, and suggest fixes. LLMs trained on historical bug data can enable prediction of likely failure points in generated code. Similarly, AI agents are used to perform static code analysis<\/a>, identify security vulnerabilities<\/a>, suggest performance improvements and ensure adherence to coding standards<\/a> and best practices.[1]<\/a><\/sup><\/p>\n\n\n\n

Beyond detection, researchers have explored using LLMs for automated program repair, where models propose candidate patches for buggy code. Off-the-shelf LLMs have been reported to repair some security-relevant defects in a zero-shot setting (i.e., without task-specific fine-tuning), including issues categorized by the\u00a0Common Weakness Enumeration<\/a>\u00a0(CWE),\u00a0being comparable to contemporary, non-AI bug fixing tools. These approaches build on LLMs\u2019 code-generation capability and the resulting patches still require validation through\u00a0software testing<\/a>,\u00a0static program analysis<\/a>, and human\u00a0code review<\/a><\/p>\n\n\n\n

Challenges<\/h6>\n\n\n\n

The incorporation of AI tools has introduced new\u00a0ethical dilemmas<\/a>\u00a0and\u00a0intellectual property<\/a>\u00a0challenges. The ownership of AI-generated code is unclear: who is responsible for the generated end-product? Also unclear are the ethical responsibilities of generated code.Changes in the role of\u00a0software engineers<\/a>\u00a0are inevitable.<\/p>\n\n\n\n

Governance and oversight<\/h6>\n\n\n\n

The outputs from AI-assisted software development require to be validated through a combination of automated testing, static analysis tools and human review, creating a\u00a0governance layer<\/a>\u00a0that acts as a safeguard ensuring quality and accountability.<\/p>\n\n\n\n

Industry perspectives<\/h6>\n\n\n\n

Technology sector leaders have highlighted the transformative potential of AI-assisted software development. In an ‘Unlocking AI Potential’ session of ‘Advancing AI 2025’ hosted by\u00a0AMD<\/a>\u00a0Developer Central,\u00a0Andrew Ng<\/a>\u00a0and\u00a0Lisa Su<\/a>\u00a0emphasized the strategic and operational implications of integrating AI tools into development workflows. Ng noted that AI systems are increasingly capable of “helping programmers focus on higher-level problem solving”, while Su framed the shift as “an opportunity to redefine performance and productivity across industries.<\/p>\n\n\n\n

Vibe coding<\/h6>\n\n\n\n

In computer programming<\/a>, vibe coding<\/a> is a software development<\/a> practice assisted by artificial intelligence (AI) such as by chatbots<\/a> (programs that simulate conversation) or AI agents<\/a> such as Codex<\/a> or Claude Code<\/a>. The software<\/a> developer describes a project or task in a prompt<\/a> to a large language model<\/a> (LLM), which generates source code<\/a> automatically. Vibe coding may involve accepting AI-generated code without reviewing the output thoroughly, instead relying on results and follow-up prompts to guide changes.[12]<\/a><\/sup>[13]<\/a><\/sup><\/p>\n\n\n\n

The term was coined by computer scientist\u00a0Andrej Karpathy<\/a>, a co-founder of\u00a0OpenAI<\/a>\u00a0and former AI leader at\u00a0Tesla<\/a>, in February 2025.\u00a0Merriam-Webster<\/a>\u00a0listed the term in March 2025 as a “slang & trending” expression.\u00a0It was named the\u00a0Collins English Dictionary<\/a><\/em>\u00a0Word of the Year<\/a>\u00a0for 2025.<\/p>\n\n\n\n

Advocates of vibe coding say that it allows even\u00a0amateur programmers<\/a>\u00a0to produce software without the extensive training and skills required for\u00a0software engineering<\/a>.[17<\/a><\/sup>[18]<\/a><\/sup>\u00a0Critics point out a lack of accountability, maintainability, and the increased risk of introducing\u00a0security vulnerabilities<\/a>\u00a0in the resulting software<\/p>\n","protected":false},"excerpt":{"rendered":"

AI-assisted software development\u00a0is the use of\u00a0artificial intelligence\u00a0(AI) to augment\u00a0software development. It uses\u00a0large language models\u00a0(LLMs),\u00a0natural language processing,\u00a0intelligent agents\u00a0and other AI technologies to assist\u00a0software developers. It helps in a range of tasks of the\u00a0software development life cycle, from initial\u00a0code generation\u00a0to subsequent\u00a0debugging,\u00a0testing\u00a0and\u00a0documentation.[1]\u00a0Agentic coding\u00a0denotes the use of more\u00a0autonomous\u00a0AI agents\u00a0for software development Technologies Code generation LLMs that have been trained […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":true,"template":"","format":"standard","meta":{"footnotes":""},"categories":[18],"tags":[21,23,25,20,19],"class_list":["post-492","post","type-post","status-publish","format-standard","hentry","category-computing","tag-bury","tag-greater-manchester","tag-radcliffe","tag-web-design","tag-web-development"],"_links":{"self":[{"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/posts\/492","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/comments?post=492"}],"version-history":[{"count":0,"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/posts\/492\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/media?parent=492"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/categories?post=492"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.derniercriweb.co.uk\/bury-blog\/wp-json\/wp\/v2\/tags?post=492"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}