{"id":5436,"date":"2026-07-20T20:19:05","date_gmt":"2026-07-20T17:19:05","guid":{"rendered":"https:\/\/lasoft.org\/blog\/?p=5436"},"modified":"2026-07-20T20:46:27","modified_gmt":"2026-07-20T17:46:27","slug":"ai-in-software-development-survey","status":"publish","type":"post","link":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/","title":{"rendered":"How We Use AI in Software Development (2026 Survey of 30 Engineers)"},"content":{"rendered":"<p>Over the past few months, we surveyed all 30 members of our delivery team, 30 people across 7 departments, about exactly how we use AI in software development. Developers, QA engineers, product managers, designers, DevOps, and business analysts. We weren\u2019t looking for hype. We wanted honest answers about where AI saves time, where it struggles, and where human expertise remains irreplaceable.<\/p>\n<p>What came back is the clearest answer we\u2019ve ever had to a question many people are Googling right now: how do software developers actually use AI tools? People use AI in production, for client work, and to meet deadlines.<\/p>\n<p><strong>Every one of our 30 team members uses AI. Every one of them still reviews its output before it reaches a product. The average productivity gain is about 30%.<\/strong><\/p>\n<h2>How Do Software Developers Use AI Tools?<\/h2>\n<p>Software developers use AI tools mainly to accelerate well-defined, repetitive, or research-heavy work: writing tests, generating scripts and migrations, analyzing unfamiliar code, drafting documentation, and writing SQL, while keeping architecture, complex business logic, and final quality review in human hands. Across our team, the use of AI tools produced an average productivity gain of about 30%, and 100% of developers reviewed AI output before merging it.<\/p>\n<h2>The Numbers We Actually Measured<\/h2>\n<p>Vague claims about tenfold productivity are precisely what we\u2019re trying to avoid, so let\u2019s be specific about what the survey found.<\/p>\n<table>\n<thead>\n<tr>\n<th><strong>Metric<\/strong><\/th>\n<th><strong>Value<\/strong><\/th>\n<th><strong>Note<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>People surveyed<\/td>\n<td>30<\/td>\n<td>All gave detailed feedback<\/td>\n<\/tr>\n<tr>\n<td>Departments covered<\/td>\n<td>7<\/td>\n<td>Development, QA, PM\/PO, DevOps, Design, BA, CRM PM<\/td>\n<\/tr>\n<tr>\n<td>Average productivity gain<\/td>\n<td>~30%<\/td>\n<td>~0% to 5x by department; up to 12x on specific tasks<\/td>\n<\/tr>\n<tr>\n<td>Team members who verify AI output<\/td>\n<td>100%<\/td>\n<td>Nobody ships AI-generated work without reviewing it<\/td>\n<\/tr>\n<tr>\n<td>Using paid tiers<\/td>\n<td>~60%<\/td>\n<td>Some pay out of pocket<\/td>\n<\/tr>\n<tr>\n<td>Most-used tool<\/td>\n<td>ChatGPT<\/td>\n<td>Present in all 7 departments<\/td>\n<\/tr>\n<tr>\n<td>Biggest single win<\/td>\n<td>Writing tests<\/td>\n<td>4x\u201312x faster<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>AI Tools Our Team Uses<\/h2>\n<p>We\u2019re deliberately not loyal to a single tool. Different models really are better at different jobs, and pretending one wins everything is how you end up with mediocre output across the board. Here\u2019s the actual stack, straight from the survey.<\/p>\n<table>\n<thead>\n<tr>\n<th><strong>Tool<\/strong><\/th>\n<th><strong>Departments<\/strong><\/th>\n<th><strong>What we use it for<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>ChatGPT<\/td>\n<td>Development, QA, PM\/PO, DevOps, Design, BA, CRM PM<\/td>\n<td>General assistant, broad tasks, working with prior context<\/td>\n<\/tr>\n<tr>\n<td>Claude \/ Claude Code<\/td>\n<td>Development, Design<\/td>\n<td>Code, legacy analysis, MCP integrations \u2014 best at understanding existing code<\/td>\n<\/tr>\n<tr>\n<td>Cursor<\/td>\n<td>Development, DevOps<\/td>\n<td>IDE agent, code review inside the editor<\/td>\n<\/tr>\n<tr>\n<td>Gemini<\/td>\n<td>Development, BA, Design, DevOps<\/td>\n<td>Research and reasoning over large data<\/td>\n<\/tr>\n<tr>\n<td>Codex (OpenAI)<\/td>\n<td>Development<\/td>\n<td>IDE integration, commits<\/td>\n<\/tr>\n<tr>\n<td>Lovable<\/td>\n<td>Design, PO<\/td>\n<td>Clickable prototypes<\/td>\n<\/tr>\n<tr>\n<td>Figma Make \/ Figma AI<\/td>\n<td>Design<\/td>\n<td>Prototypes, design-system work<\/td>\n<\/tr>\n<tr>\n<td>Claude Code + Figma MCP<\/td>\n<td>Design<\/td>\n<td>Bulk design-system changes (tokens, components)<\/td>\n<\/tr>\n<tr>\n<td>Jira AI \/ Atlassian Rovo<\/td>\n<td>PM, BA, QA<\/td>\n<td>Workflows, sprint summaries<\/td>\n<\/tr>\n<tr>\n<td>OpenAI API + MCP<\/td>\n<td>Development<\/td>\n<td>Custom pipelines, auto-commits, browser automation<\/td>\n<\/tr>\n<tr>\n<td>opencode CLI<\/td>\n<td>DevOps<\/td>\n<td>Configs, scripts<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\t\t<picture>\n\t\t\t<source\n\t\t\t\tmedia=\"(max-width: 424px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=425 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=850 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1275 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(max-width: 424px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=425 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=850 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=1275 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 425px) and (max-width: 479px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=480 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=960 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1440 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 425px) and (max-width: 479px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=480 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=960 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=1440 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=2304 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=2304 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;w=2304 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 768px) and (max-width: 1023px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=464 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=928 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1392 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 768px) and (max-width: 1023px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=464 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=928 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1392 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1024px) and (max-width: 1199px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=552 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1104 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1656 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1200px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=624 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1248 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1872 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1200px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=624 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1248 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=1872 3x\">\n\n\t\t\t<img\n\t\t\t\tsrc=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2Fhow-we-use-ai-in-software-development.png&#038;output=webp&#038;w=624\"\n\t\t\t\talt=\"how we use ai in software development\"\n\t\t\t\tdecoding=\"async\"\n\t\t\t\tloading=\"lazy\"\n\t\t\t\twidth=\"100%\">\n\t\t<\/picture>\n<\/p>\n<p>In practice, this is how we choose between AI tools:<\/p>\n<ul>\n<li>Gemini for research and analysis over large data sets.<\/li>\n<li>ChatGPT for broad tasks and anything that builds on earlier context.<\/li>\n<li>Claude for code, legacy analysis, and MCP integrations, as it reads unfamiliar codebases best.<\/li>\n<li>Cursor is useful, but we deliberately limit its scope.<\/li>\n<li>Paid tiers are preferred because they yield materially better results.<\/li>\n<\/ul>\n<p>That last one came through loudly. Free tiers produce noticeably worse output, including flat-out errors in things like formulas. About 60% of the team pays for premium plans.<\/p>\n<h2>Where AI Coding Tools Genuinely Shine<\/h2>\n<p>The averages hide the interesting part. Gains are wildly uneven, not just between departments but between two developers sitting side by side. Here\u2019s where the value actually showed up.<\/p>\n<h3>Development (14 people)<\/h3>\n<table>\n<thead>\n<tr>\n<th><strong>Task<\/strong><\/th>\n<th><strong>Before AI<\/strong><\/th>\n<th><strong>After AI<\/strong><\/th>\n<th><strong>Speed-up<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Unit tests<\/td>\n<td>~1 hour<\/td>\n<td>5\u201315 min<\/td>\n<td>4x\u201312x<\/td>\n<\/tr>\n<tr>\n<td>Scripts \/ RegExp \/ migrations<\/td>\n<td>~1 hour<\/td>\n<td>~5 min<\/td>\n<td>up to 12x<\/td>\n<\/tr>\n<tr>\n<td>Integration tests for an endpoint<\/td>\n<td>~1 day<\/td>\n<td>40\u201360 min<\/td>\n<td>4x\u20136x<\/td>\n<\/tr>\n<tr>\n<td>Analyzing legacy \/ unfamiliar code<\/td>\n<td>1\u20132 hours<\/td>\n<td>20\u201330 min<\/td>\n<td>3x\u20135x<\/td>\n<\/tr>\n<tr>\n<td>Typical feature (component + API + types)<\/td>\n<td>2\u20133 hours<\/td>\n<td>40\u201360 min<\/td>\n<td>3x\u20134x<\/td>\n<\/tr>\n<tr>\n<td>SQL problem under load<\/td>\n<td>3\u20134 days<\/td>\n<td>1 hour<\/td>\n<td>10x+<\/td>\n<\/tr>\n<tr>\n<td>Complex CRM logic<\/td>\n<td>1 sprint<\/td>\n<td>2 days<\/td>\n<td>5x+<\/td>\n<\/tr>\n<tr>\n<td>Code documentation<\/td>\n<td>baseline<\/td>\n<td>2\u20133x faster<\/td>\n<td>2x\u20133x<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>These are real tickets. A SQL performance problem that once took three or four days was resolved in about an hour. Migration scripts that took an hour are done in five minutes. That\u2019s not hype; it\u2019s the kind of grinding, well-scoped work AI is genuinely great at.<\/p>\n<p>But the same developers were just as specific about where AI actively hurts if you let it run unsupervised:<\/p>\n<ul>\n<li><strong>\u201cSolve it at any cost.\u201d<\/strong> It often produces code that works but ignores better architectural decisions.<\/li>\n<li><strong>Redundant, duplicated code.<\/strong> It rewrites existing validation and ignores existing logic.<\/li>\n<li><strong>Scope creep.<\/strong> It edits unrelated functions or files it was never asked to touch.<\/li>\n<li><strong>Recursive \u201cfixing.\u201d<\/strong> Fix\/test loops with no real progress \u2014 one dev watched it spin for 3+ iterations.<\/li>\n<li><strong>Cursor touching too much.<\/strong> Hard to track the diff on a 100-file commit.<\/li>\n<li><strong>Outdated libraries.<\/strong> It references deprecated methods even when you name the version.<\/li>\n<li><strong>Long-chat decay.<\/strong> It loses the thread after enough messages.<\/li>\n<\/ul>\n<p>Read that list again and notice something: none of it is a reason to stop using AI. Every item is a reason you need an experienced engineer reviewing what it produces. Which is the whole argument.<\/p>\n<h3>PM \/ PO<\/h3>\n<p>Operational work turned out to be a near-perfect fit.<\/p>\n<table>\n<thead>\n<tr>\n<th><strong>Task<\/strong><\/th>\n<th><strong>Before AI<\/strong><\/th>\n<th><strong>After AI<\/strong><\/th>\n<th><strong>Gain<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Writing tickets (US\/AC)<\/td>\n<td>30\u201360 min<\/td>\n<td>15\u201325 min<\/td>\n<td>50\u201360%<\/td>\n<\/tr>\n<tr>\n<td>Documentation<\/td>\n<td>2\u20133 hours<\/td>\n<td>1\u20131.5 hours<\/td>\n<td>40\u201350%<\/td>\n<\/tr>\n<tr>\n<td>Follow-ups \/ communication<\/td>\n<td>20\u201330 min<\/td>\n<td>5\u201310 min<\/td>\n<td>60\u201375%<\/td>\n<\/tr>\n<tr>\n<td>Manual reports<\/td>\n<td>2\u20133 hours<\/td>\n<td>up to 1 hour<\/td>\n<td>50\u201365%<\/td>\n<\/tr>\n<tr>\n<td>Risk management<\/td>\n<td>manual<\/td>\n<td>structured list + mitigation plans<\/td>\n<td>significant<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n<table>\n<tbody>\n<tr>\n<td>QA<\/td>\n<td>QA is where the survey got refreshingly unhyped. The team uses AI mostly as an assistant for test documentation and technical artifacts: generating checklists; working through SQL, logs, DOM elements, and DevTools; and planning load tests.\n<p>The productivity spread aligns almost perfectly with engagement. Deep integration into technical work (code, logs, SQL) delivered the top gains, around +30%. Lighter use landed at +10\u201320%. And for a few skeptical team members, AI was actually a drag, an extra step that meant rewriting prompts and double-checking output and slowed them down. We think reporting that honestly matters more than pretending everyone won.<\/p><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n<table>\n<tbody>\n<tr>\n<td>Design\n<\/td><td>Clickable prototypes for validating an idea, built in Lovable or Figma Make, now take about two days. Illustration generation reduces time by roughly 40% compared to hand-drawn illustrations. Bulk design-system changes through Claude Code + Figma MCP save 20\u201330%. Same caveat as everywhere else: designers said AI \u201cignores the specific request and adds extra,\u201d and about 60% of prompts land in 1\u20133 attempts, while the rest need to be broken down and reformulated. It structures and suggests. It doesn\u2019t decide.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n<table>\n<tbody>\n<tr>\n<td>DevOps\n<p>BA<\/p><\/td>\n<td>DevOps rated AI excellent for writing and checking scripts and for migrating CI\/CD between platforms (say, CircleCI \u2192 GitHub Actions); good for K8s configs and security checks; and weak for Terraform, OIDC migrations, and, memorably, hosting-price math (\u201calways gets it wrong\u201d). One engineer\u2019s incident habit is the ideal mental model: hand the logs to the AI agent and dig into them yourself in parallel; the different perspectives often help.\n<p>Business analysts got their best results from structured chats scoped to a single role \u2013 BA helper here, a copywriter there, a proofreader. Deep research compressed days of forum and video hunting into 2\u20135 hours.<\/p><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>\t\t<picture>\n\t\t\t<source\n\t\t\t\tmedia=\"(max-width: 424px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=425 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=850 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1275 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(max-width: 424px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=425 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=850 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=1275 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 425px) and (max-width: 479px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=480 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=960 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1440 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 425px) and (max-width: 479px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=480 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=960 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=1440 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=2304 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=2304 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 480px) and (max-width: 767px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=768 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=1536 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;w=2304 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 768px) and (max-width: 1023px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=464 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=928 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1392 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 768px) and (max-width: 1023px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=464 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=928 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1392 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1024px) and (max-width: 1199px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=552 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1104 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1656 3x\">\n\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1200px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=624 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1248 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1872 3x\"\n\t\t\t\ttype=\"image\/webp\">\n\t\t\t<source\n\t\t\t\tmedia=\"(min-width: 1200px)\"\n\t\t\t\tsrcset=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=624 1x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1248 2x,\n                \t\t\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=1872 3x\">\n\n\t\t\t<img\n\t\t\t\tsrc=\"\/\/wsrv.nl\/?url=https%3A%2F%2Flasoft.org%2Fblog%2Fwp-content%2Fuploads%2F2026%2F07%2FAI-Assisted-Software-Development-Process.png&#038;output=webp&#038;w=624\"\n\t\t\t\talt=\"how we use ai in software development\"\n\t\t\t\tdecoding=\"async\"\n\t\t\t\tloading=\"lazy\"\n\t\t\t\twidth=\"100%\">\n\t\t<\/picture>\n<\/h2>\n<h2>The Part Most Vendors Leave Out: Where AI Fails<\/h2>\n<p>If a team only shows you the wins, that\u2019s marketing, not reporting. Here are the complaints the whole company raised, ranked by severity.<\/p>\n<table>\n<thead>\n<tr>\n<th><strong>Problem<\/strong><\/th>\n<th><strong>Departments<\/strong><\/th>\n<th><strong>Severity<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>No context of the actual system<\/td>\n<td>All (esp. QA \/ online chats)<\/td>\n<td>\u2605\u2605\u2605\u2605 Critical<\/td>\n<\/tr>\n<tr>\n<td>Mandatory human review<\/td>\n<td>All 30 people<\/td>\n<td>\u2605\u2605\u2605\u2605 Always<\/td>\n<\/tr>\n<tr>\n<td>Hallucinations \/ non-objective conclusions<\/td>\n<td>Dev, QA, BA<\/td>\n<td>\u2605\u2605\u2605 High<\/td>\n<\/tr>\n<tr>\n<td>Long-chat degradation<\/td>\n<td>All<\/td>\n<td>\u2605\u2605\u2605 High<\/td>\n<\/tr>\n<tr>\n<td>Outdated libraries and methods<\/td>\n<td>Dev, QA<\/td>\n<td>\u2605\u2605 Medium<\/td>\n<\/tr>\n<tr>\n<td>Redundant \/ duplicated code<\/td>\n<td>Development<\/td>\n<td>\u2605\u2605 Medium<\/td>\n<\/tr>\n<tr>\n<td>Changes outside task scope<\/td>\n<td>Development (Cursor)<\/td>\n<td>\u2605\u2605 Medium<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>AI tools are APIs. The processing parameters can change without warning. The model can get worse the moment a new version ships. The token algorithm can change. Your build\u2019s behavior and its cost can shift because an AI provider two steps removed from you pushed an update overnight. Predictability is genuinely in question, which is why we treat AI as a powerful assistant within a controlled process, never as the autonomous author of your product.<\/p>\n<p>The most dangerous version of this scenario shows up when you lean on <a href=\"https:\/\/lasoft.org\/blog\/how-to-build-an-ai-agent-for-business\/\">AI agents<\/a> instead of assistants: the design and detailed-verification layer can quietly vanish. Our stance, shaped directly by this survey, is simple: AI belongs in the hands of an engineer who owns the outcome.<\/p>\n<h2>How AI Changed Software Development Work<\/h2>\n<p>Here\u2019s the finding that reframes the whole conversation. AI didn\u2019t just make our people faster at their existing jobs. It changed what the job is.<\/p>\n<table>\n<thead>\n<tr>\n<th><strong>Before AI<\/strong><\/th>\n<th><strong>With AI<\/strong><\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Writing lines of code<\/td>\n<td>Reviewing and quality-controlling code<\/td>\n<\/tr>\n<tr>\n<td>Searching forums and docs<\/td>\n<td>Formulating a precise prompt<\/td>\n<\/tr>\n<tr>\n<td>Tests as a separate, \u201clater\u201d task<\/td>\n<td>Tests written as part of the main work<\/td>\n<\/tr>\n<tr>\n<td>Documentation always deferred<\/td>\n<td>Documentation produced in parallel<\/td>\n<\/tr>\n<tr>\n<td>Long onboarding into new tech<\/td>\n<td>Fast start with AI support<\/td>\n<\/tr>\n<tr>\n<td>Architecture as a side task<\/td>\n<td>Architecture as the human\u2019s core focus<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>That\u2019s the line the whole team keeps landing on, the consensus of all 30 people:<\/p>\n<p><strong>AI replaces \u201cwriting lines of code and text.\u201d It does not replace thinking, responsibility, or engineering judgment.<\/strong><\/p>\n<p>The work moves up the stack toward architecture, review, and decisions. The barrier to entry for new technologies (Playwright, TypeScript, SQL for QA) has dropped significantly. But knowing the language and the system got more important, not less, because that\u2019s what lets you review and debug well.<\/p>\n<div class=\"laTeaser\">\n<div class=\"laTeaser__content laTeaser__light\">\n<div class=\"laTeaser__img\"><\/div>\n<div class=\"laTeaser__txt\">\n<h3 class=\"laTeaser__h3\">Curious how an AI-assisted engineering team would approach your product?<\/h3>\n<p>We\u2019ll show you exactly where AI accelerates delivery, where human expertise matters most, and how we balance both to build reliable software.<\/p>\n<div class=\"laTeaser__lnk\"><a href=\"https:\/\/lasoft.org\/#contactForm\">Let\u2019s talk<\/a><\/div>\n<\/div>\n<\/div>\n<\/div>\n<h2>How We Use AI Responsibly: Our Working Practices<\/h2>\n<p>\u201cWe use AI responsibly\u201d means nothing without specifics, so here are the actual habits the team converged on.<\/p>\n<p><strong>Prompting<\/strong><\/p>\n<ul>\n<li>Be explicit: \u201cdo X,\u201d \u201cdon\u2019t do Y\u201d \u2014 concrete constraints in every request.<\/li>\n<li>Give full task context plus access to the existing code.<\/li>\n<li>Have the model ask when something\u2019s unclear rather than invent an answer.<\/li>\n<li>One chat, one role, one topic. Start fresh after 3\u20134 exchanges to avoid context rot.<\/li>\n<\/ul>\n<p><strong>Workflow<\/strong><\/p>\n<ul>\n<li>Start with the overall context, then break the work into small pieces.<\/li>\n<li>TDD with AI: AI writes the tests, a human refactors, and AI generates the implementation.<\/li>\n<li>Small commits, so rollback and review stay easy.<\/li>\n<li>Pipeline across models: each AI does what it\u2019s best at; the human synthesizes.<\/li>\n<li>During incidents, AI and humans investigate in parallel.<\/li>\n<li>Cross-check one model against another (Gemini \u2194 ChatGPT).<\/li>\n<\/ul>\n<p><strong>Tool choice<\/strong><\/p>\n<ul>\n<li>Match the tool to the task (see the routing guide above).<\/li>\n<li>Keep Cursor\u2019s change scope under control.<\/li>\n<li>Pay for the premium tiers.<\/li>\n<\/ul>\n<h2>Why This Matters When You Choose Who Builds Your Software<\/h2>\n<p>We\u2019ve audited vibe-coded SaaS products and found nine critical issues, reviewed <a href=\"https:\/\/lasoft.org\/blog\/is-lovable-right-for-your-project-a-technical-review-from-prototype-to-production\/\">Lovable from prototype to production<\/a>, and helped founders stabilize systems that AI built but couldn\u2019t scale. The through-line never changes: AI is a multiplier for people who know how to steer it and a trap for people who trust it without checking.<\/p>\n<p>That\u2019s not an argument against an AI-assisted vendor. It\u2019s an argument for asking harder questions. Before you pick who builds your product, ask:<\/p>\n<ul>\n<li>Which tools do you use, and for which tasks? A real answer is specific, like the table above.<\/li>\n<li>Who reviews AI-generated code before it ships? The only acceptable answer is a qualified human, every time.<\/li>\n<li>How do you handle the things AI is bad at: architecture, complex business logic, and scope control?<\/li>\n<li>What\u2019s your policy on autonomous agents versus assisted development?<\/li>\n<\/ul>\n<p>At LaSoft, the answers are built into how we work \u2014 and this survey is the evidence. Thirty people use AI every day. Thirty people review its output every time. The speed is real, about 30% on average and far more on the right tasks, but the responsibility never leaves human hands. AI changed how we build software. It didn\u2019t change who\u2019s accountable for it.<\/p>\n<h2>FAQs<\/h2>\n<details>\n<summary>\n<h4>How do IT experts use AI in software development in 2026?<\/h4>\n<\/summary>\n<div>Developers use AI tools to speed up well-defined, repetitive work, such as writing unit and integration tests, generating scripts and database migrations, analyzing legacy code, drafting documentation, and writing SQL, while keeping architecture, complex business logic, and final review in human hands.<\/div>\n<\/details>\n<details>\n<summary>\n<h4>Do developers still review AI-generated code?<\/h4>\n<\/summary>\n<div>In our survey, 100% of team members reviewed AI output before it reached a product. AI frequently produces working-but-suboptimal code, duplicates existing logic, edits files outside the task scope, and references outdated libraries, all of which require a qualified engineer to catch. Human review isn\u2019t a recommendation; for us it\u2019s a rule.<\/div>\n<\/details>\n<details>\n<summary>\n<h4>What are the best AI coding tools for software development?<\/h4>\n<\/summary>\n<div>There\u2019s no single winner; it rather depends on the task. In our team, Claude and Claude Code lead in understanding and analyzing code; ChatGPT is the general-purpose workhorse; Gemini is strongest for research over large data; and Cursor is the preferred in-editor agent (kept on a tight scope). Paid tiers consistently outperform free ones.<\/div>\n<\/details>\n<details>\n<summary>\n<h4>What can\u2019t AI do in software development?<\/h4>\n<\/summary>\n<div>AI struggles with anything requiring deep context about your specific system, sound architectural judgment, complex or safety-critical business logic, and reliable factual precision (it hallucinates and cites outdated methods). It also degrades over long chat sessions. These are precisely the areas where experienced engineers remain essential. <\/div>\n<\/details>\n<details>\n<summary>\n<h4>Is it safe to build a product entirely with AI or with <a href=\"https:\/\/lasoft.org\/blog\/we-audited-a-vibe-coded-saas-product-and-found-9-critical-issues\/\">vibe coding<\/a>?<\/h4>\n<\/summary>\n<div>Building a full product with AI alone is risky. Autonomous AI agents tend to skip the design and verification layers, leading to unpredictable performance issues and bugs that surface later, often when the system needs to scale. AI works best as an assistant inside a controlled process owned by a human engineer, not as the sole author of production software.<\/div>\n<\/details>\n<details>\n<summary>\n<h4>How can I tell if a software vendor uses coding AI tools responsibly?<\/h4>\n<\/summary>\n<div>Ask specific questions: which tools they use and for which tasks, who reviews AI-generated code before it ships, how they handle work that AI is bad at, such as architecture, complex logic, and scope control, and what their policy is on autonomous agents versus assisted development. A responsible vendor answers concretely and can point to a human reviewing every piece of AI output.<\/div>\n<\/details>\n","protected":false},"excerpt":{"rendered":"Over the past few months, we surveyed all 30 members of our delivery team, 30 people across 7 departments, about exactly how we use AI in software development. Developers, QA engineers, product managers, designers, DevOps, and business analysts. We weren\u2019t looking for hype. We wanted honest answers about where AI saves time, where it struggles,&hellip;","protected":false},"author":20,"featured_media":5446,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[180],"tags":[],"coauthors":[185],"class_list":["post-5436","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-software-development"],"yoast_head":"<title>How We Use AI in Software Development (2026 Data)<\/title>\n<meta name=\"description\" content=\"We surveyed 30 engineers on how they really use AI tools. The honest numbers, tools, and limits of AI-assisted software development.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"How We Use AI in Software Development (2026 Data)\" \/>\n<meta property=\"og:description\" content=\"We surveyed 30 engineers on how they really use AI tools. The honest numbers, tools, and limits of AI-assisted software development.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/\" \/>\n<meta property=\"og:site_name\" content=\"Software Development Blog\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/lasoftians\/\" \/>\n<meta property=\"article:published_time\" content=\"2026-07-20T17:19:05+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-07-20T17:46:27+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1344\" \/>\n\t<meta property=\"og:image:height\" content=\"768\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Alena Porokh\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@LaSoftAgency\" \/>\n<meta name=\"twitter:site\" content=\"@LaSoftAgency\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Alena Porokh\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"12 minutes\" \/>","yoast_head_json":{"title":"How We Use AI in Software Development (2026 Data)","description":"We surveyed 30 engineers on how they really use AI tools. The honest numbers, tools, and limits of AI-assisted software development.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/","og_locale":"en_US","og_type":"article","og_title":"How We Use AI in Software Development (2026 Data)","og_description":"We surveyed 30 engineers on how they really use AI tools. The honest numbers, tools, and limits of AI-assisted software development.","og_url":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/","og_site_name":"Software Development Blog","article_publisher":"https:\/\/www.facebook.com\/lasoftians\/","article_published_time":"2026-07-20T17:19:05+00:00","article_modified_time":"2026-07-20T17:46:27+00:00","og_image":[{"width":1344,"height":768,"url":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png","type":"image\/png"}],"author":"Alena Porokh","twitter_card":"summary_large_image","twitter_creator":"@LaSoftAgency","twitter_site":"@LaSoftAgency","twitter_misc":{"Written by":"Alena Porokh","Est. reading time":"12 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/#article","isPartOf":{"@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/"},"author":{"name":"Alena Porokh","@id":"https:\/\/lasoft.org\/blog\/#\/schema\/person\/a11f7419682a192db21576b723c30eb4"},"headline":"How We Use AI in Software Development (2026 Survey of 30 Engineers)","datePublished":"2026-07-20T17:19:05+00:00","dateModified":"2026-07-20T17:46:27+00:00","mainEntityOfPage":{"@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/"},"wordCount":2368,"publisher":{"@id":"https:\/\/lasoft.org\/blog\/#organization"},"image":{"@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/#primaryimage"},"thumbnailUrl":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png","articleSection":["Software Development"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/","url":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/","name":"How We Use AI in Software Development (2026 Data)","isPartOf":{"@id":"https:\/\/lasoft.org\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/#primaryimage"},"image":{"@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/#primaryimage"},"thumbnailUrl":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png","datePublished":"2026-07-20T17:19:05+00:00","dateModified":"2026-07-20T17:46:27+00:00","description":"We surveyed 30 engineers on how they really use AI tools. The honest numbers, tools, and limits of AI-assisted software development.","inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/lasoft.org\/blog\/ai-in-software-development-survey\/#primaryimage","url":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png","contentUrl":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2026\/07\/how-we-use-ai-in-software-development-2.png","width":1344,"height":768,"caption":"how we use ai coding tools in lasoft"},{"@type":"WebSite","@id":"https:\/\/lasoft.org\/blog\/#website","url":"https:\/\/lasoft.org\/blog\/","name":"Software Development Blog","description":"Insightful Analysis of IT Markets and Emerging Technologies","publisher":{"@id":"https:\/\/lasoft.org\/blog\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/lasoft.org\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/lasoft.org\/blog\/#organization","name":"LaSoft","url":"https:\/\/lasoft.org\/blog\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/lasoft.org\/blog\/#\/schema\/logo\/image\/","url":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2024\/06\/lasoft.jpg","contentUrl":"https:\/\/lasoft.org\/blog\/wp-content\/uploads\/2024\/06\/lasoft.jpg","width":1200,"height":628,"caption":"LaSoft"},"image":{"@id":"https:\/\/lasoft.org\/blog\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/lasoftians\/","https:\/\/x.com\/LaSoftAgency","https:\/\/www.instagram.com\/lasoftians\/","https:\/\/www.linkedin.com\/company\/la'soft-","https:\/\/www.youtube.com\/channel\/UCGiINdBnqPoBpK1WaCEZMHQ\/"]},{"@type":"Person","@id":"https:\/\/lasoft.org\/blog\/#\/schema\/person\/a11f7419682a192db21576b723c30eb4","name":"Alena Porokh","description":"Alena Porokh, as a member of the LaSoft Marketing team, focuses on creating professional content for the company's blog to address the business development issues of LaSoft\u2019s target audience. With over 5 years of experience as a Content Manager in the software development and video game development industries, Alena plays a pivotal role in our content strategy. Her goal is to offer followers and potential clients brilliant business ideas and identify growth points for their commercial success. Her expertise and dedication ensure that LaSoft\u2019s content is both engaging and effective in driving business growth.","sameAs":["https:\/\/linkedin.com\/in\/alenaporokh\/"],"url":"https:\/\/lasoft.org\/blog\/author\/porokh\/"}]}},"_links":{"self":[{"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/posts\/5436","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/users\/20"}],"replies":[{"embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/comments?post=5436"}],"version-history":[{"count":15,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/posts\/5436\/revisions"}],"predecessor-version":[{"id":5454,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/posts\/5436\/revisions\/5454"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/media\/5446"}],"wp:attachment":[{"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/media?parent=5436"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/categories?post=5436"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/tags?post=5436"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/lasoft.org\/blog\/wp-json\/wp\/v2\/coauthors?post=5436"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}