State of vibecoding in Feb 2026 (summary version)
NOTE: Prefer a more in-depth version of this paper? Check the mad scientist version here.
Summary
Vibe coding describes the practice of building software primarily through natural-language interaction with AI systems. Rather than writing code line by line, humans describe intent and evaluate results while AI performs implementation.
In 2025, this shifted from experimental workflow to mainstream practice. Adoption accelerated rapidly across startups and major enterprises. While productivity gains are real, research has also revealed elevated security and quality risks.
For business leaders, the strategic question is not whether teams are using AI in development — they are. The question is how to govern, secure, and architect responsibly in an AI-accelerated environment.
Acceleration of Software Creation/ Custom Code
Adoption data from 2025 illustrates the scale of change:
25% of Y Combinator’s Winter 2025 startups reported codebases that were 95% AI-generated. [3]
84% of developers reported using or planning to use AI tools in their workflow. [4]
Google disclosed that over 25% of its new code was generated by AI and reviewed by engineers. [5]
iOS app releases increased roughly 60% year-over-year in late 2025. [6]
The cost and time required to prototype internal tools or new products has dropped significantly.
Risk and Security Findings
Independent research has highlighted structural risks:
Experienced developers were 19% slower using AI tools on familiar repositories despite perceiving gains. [7]
Approximately 45% of AI-generated code samples introduced common OWASP risk vulnerabilities. [8]
A security review across major vibe coding tools found 69 vulnerabilities across 15 test applications. [9]
AI co-authored pull requests showed 2.74× higher rates of security vulnerabilities in one large analysis. [10]
Public incidents demonstrated AI agents making destructive operational errors. [11]
The acceleration of production has not automatically produced secure or maintainable systems.
Strategic Implications for Leadership
AI-assisted coding is already embedded in development workflows. The competitive advantage will not come from merely adopting AI tools, but from governing them effectively.
Leaders must prioritize:
Architectural oversight over all technology created by your organization
Security governance
Risk management controls
Clear product intent and constraint setting
Ongoing system stewardship
Vibe coding lowers the barrier to generating software. It does not eliminate the need for accountability. In fact, it ups it.
The future norm is likely to be that organizations that traditionally would never have coded anything themselves are going to start creating purpose-built custom given the ease of doing so.
Working Definition
Vibe coding is the development of software primarily through natural-language interaction with AI agents, where humans specify intent and evaluate results while AI performs implementation. This can be done by non-engineers.
The human role increasingly centers on defining systems, constraints, and evaluation criteria — not typing code.
This lowers the bar for creating software/code to solve business problems or address customer service. This opportunity is huge and will only grow more powerful as the AI-assistive tools continue to increase in quality.
Conclusion
The defining tension of 2026 is whether improved AI coding models and tooling can close the gap between:
Software that works (easy to create a "demo"-- harder to create code that is secure, scalable and reliable)
Software that is secure, maintainable, and reliable
Generating software has become dramatically easier. Governing it has not.
The opportunity for "natural language" code writing is huge -- especially for small and under resourced organizations like arts and culture organizations and nonprofits. The opportunity will only grow more powerful as the AI-assistive tools continue to increase in quality. With this opportunity comes risk. Risk of a non-codewriting organization becoming one that writes code but lacks internal governance and knowledge about safe code-writing.
*This resource paper was co-written by Claude Opus 4.6, ChatGPT 5.2, and Kristin Darrow collaboratively. Kristin Darrow is AI consultant and organizational strategist for arts and culture organizations. Find her at kristin@matters.work.
Footnotes
Merriam-Webster, “Vibe Coding” https://www.merriam-webster.com/slang/vibe-coding
Collins Dictionary, Word of the Year 2025 https://www.collinsdictionary.com/us/woty
TechCrunch, “A quarter of startups in YC’s current cohort have codebases almost entirely AI-generated” https://techcrunch.com/2025/03/06/a-quarter-of-startups-in-ycs-current-cohort-have-codebases-that-are-almost-entirely-ai-generated/
Stack Overflow 2025 Developer Survey https://survey.stackoverflow.co/2025/
The Verge, “Google says over 25% of its new code is AI-generated” https://www.theverge.com/2024/10/29/24282757/google-new-code-generated-ai-q3-2024
Business of Apps, “Agentic coding drives iOS app releases…” https://www.businessofapps.com/news/agentic-coding-drives-ios-app-releases-to-the-highest-level-ever/
METR, “Early 2025 AI Experienced OSS Developer Study” https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
Veracode, “GenAI Code Security Report” https://www.veracode.com/blog/genai-code-security-report/
CSO Online, “Output from vibe coding tools prone to critical security flaws” https://www.csoonline.com/article/4116923/output-from-vibe-coding-tools-prone-to-critical-security-flaws-study-finds.html
ITPro, “AI is creating more software flaws…” https://www.itpro.com/software/development/ai-is-creating-more-software-flaws-and-theyre-getting-worse
Fortune, “AI coding tool Replit wiped database…” https://fortune.com/2025/07/23/ai-coding-tool-replit-wiped-database-called-it-a-catastrophic-failure/
