Vibe coding is expanding beyond the realm of software development teams into a variety of business units at a range of enterprises, and technology leaders are not only supporting these efforts but in some cases leading the charge.

This democratization of software development, buoyed by vibe coding tools, including chatbots and agents that generate code via prompts, has made vibe coding a thing in departments such as human resources and marketing.

With guidance from IT, such efforts can cut down development backlogs, move solution-building closer to business problems, and open opportunities not previously pursued. But vibe coding can be edgy business. IT leaders interested in unleashing the power of vibe coding in the hands of business users must establish the governance and guardrails necessary to ensure secure results.

Here is a look at how several IT leaders are deploying vibe coding beyond IT and the challenges they’ve faced.

Feeling the vibe

At financial services technology provider EnFi, everyone in the organization, including the C-suite, is actively encouraged to use Claude Code to build their own sub-agents and then get the rest of their teams to use their creations.

“The results have surprised us,” says Scott Weller, CTO. “What started as an engineering productivity initiative has become a company-wide capability, where anyone from the CEO to a customer success manager can turn an idea into a working prototype in hours, not weeks.”

Vibe coding using the tool quickly expanded when non-engineering employees realized they could participate directly in building applications, Weller says. “Product leadership, customer success, and executive stakeholders now routinely initiate development work through the same system,” he says. “Every branch, regardless of who initiates [coding projects], passes through the same automated quality gates, architectural checks, and human code review before anything reaches production.”

When an AI agent writes code, “it follows the same rules as a senior engineer,” Weller says. “AI-powered code review catches architectural violations, security issues, and pattern inconsistencies before a human reviewer sees it. This means the output from a product manager’s Slack request meets the same quality bar as an engineer’s pull request.”

Enabling everyone in the company to come up with an idea and see it running by simply asking a bot to build it “has fundamentally changed who participates in building the product,” Weller says.

Making AI-assisted coding broadly available has shortened the time to completing some projects from weeks to hours, Weller notes. “It has dramatically increased the number of experiments we can run, on [user experience] improvements, feature explorations, and workflow alternatives,” he says.

The most significant benefit is that new development ideas are no longer bottlenecked by lack of engineering capacity. “When anyone in the company can describe what they want and see it built, the rate of experimentation goes up dramatically,” Weller says.

The momentum of building solutions to business problems

Skillsoft, a provider of technology training services and products, is also pushing vibe coding outside its software development teams.

“While it initially emerged within IT teams, we’ve been intentional about encouraging this mindset beyond traditional software development,” says Orla Daly, CIO. “We see real value in enabling teams across the organization to explore, experiment, and problem‑solve using AI in ways that are directly connected to their day‑to‑day work.”

Within Skillsoft’s infrastructure and operations organization non-development individuals are building new products and capabilities or leveraging the principles of vibe coding to troubleshoot production issues, Daly says.

“We’ve seen this show up through cross‑functional experimentation such as prototyping customer intelligence solutions for our go-to-market teams and rapid development of a customer portal,” Daly says. “When people are learning and applying AI to solve a real business problem, it creates purpose and momentum.”

With vibe coding, teams are not just learning new concepts, “they’re developing judgment, curiosity, and a better understanding of how AI can elevate their role and the business more broadly,” Daly says. “The benefits show up in very tangible ways. Teams learn faster because they’re applying AI directly to their work. Engagement increases when people feel trusted to explore and contribute ideas. And as an organization, we gain better visibility into where skills already exist and how they’re evolving.”

More broadly, vibe coding supports adaptability, Daly says. “It helps people move from seeing AI as something abstract or intimidating to something they can work with thoughtfully, using judgment and collaboration rather than relying on rigid processes,” she says. “The solutions created as a result of vibe coding have filled capability gaps and delivered solutions to production in shorter timeframes, producing real value.”

Emphasizing experimentation

At Corevist, an ecommerce platform provider, broad use of vibe coding didn’t start as a formal initiative. “It showed up in different parts of the business, mostly in sales and customer-facing teams, because people were trying to move faster and make ideas easier to communicate,” says Terry Stahler, CIO and chief customer officer.

It also didn’t spread purely on its own. “Without leadership pushing for experimentation, it likely would have moved at a much slower, consensus-driven pace,” Stahler says. “Instead, there was clear encouragement to try things, along with budget behind it.”

Today, Corevist uses vibe coding mainly as a prototyping tool. “It helps people get to something concrete faster, which makes conversations a lot clearer,” Stahler says. “That has been the biggest benefit. We are not using it as a shortcut to production software. Anything that is going to live beyond a prototype goes through our normal engineering and security process.”

Sales was the first place where Stahler saw use cases emerge. One was live prototyping during the sales process. “In a normal B2B conversation, a prospect explains what they need, the account team takes notes, and then everyone goes back and tries to interpret it later,” he says. “What changed here was the ability to turn an idea into something visible much faster, sometimes even while the conversation was still happening.”

That gave the prospect something concrete to react to and made it easier to tell whether a sales rep was actually understanding the request correctly. “It cut down on ambiguity early, which is valuable in any complex sales cycle,” Stahler says.

Marketing has also used vibe coding, for an update of its website. “The value there has been speed in the early creative and planning stages,” Stahler says. “They can generate rough versions of pages and flows much faster than they could through written direction or static mockups alone. That has made it easier to align on direction and has reduced some of the usual back-and-forth that comes with a website rebuild.”

Supplementing the engineering queue

Business growth platform provider ZenBusiness encourages vibe coding “across the board,” says Alex Victoria, CTO. “The way we think about it is pretty simple: If AI tools can help you go from idea to something working without waiting in an engineering queue, we want you to do that,” he says.

Vibe coding began within the product and engineering teams but has spread to other departments.

“We’ve seen people on the data side building their own query tools, product managers creating interactive prototypes with no design background, and teams across the company using AI to handle tasks they used to outsource to specialists,” Victoria says. “The culture we’ve tried to build is one where experimentation isn’t reserved for engineers. If you’re willing to learn and use the tools we have, anyone can code to improve their workflows on their own.”

Although Victoria has been using agentic coding tools for years, he says the “real unlock” of benefits came in 2025, with the release of Claude Code.

“I’ve seen a huge impact since then, and also with Cursor and Codex,” he says. “When someone can build their own tool or prototype in an afternoon instead of waiting weeks in an engineering queue, it changes what’s worth doing.”

The bigger benefit is how it’s changed the relationship between roles, Victoria says. “People who know how to build software can now produce far more value than ever before,” he says. “There’s still a gap between a vibe-coded idea and a running product. If you know how to ship a running product, you’re very valuable in this world.”

Facing down challenges

Embracing more widespread use of vibe coding in the enterprise comes with its own set of challenges.

One of these is maintaining quality when everyone can build. “When you open development to non-engineers, the risk is code that works but doesn’t follow project conventions, creating technical debt faster than manual development,” Weller says.

EnFi has addressed this by investing heavily in the rules and skills layer of coding. More than 40 custom skills impact AI output to match pre-determined architectural patterns.

“The agent doesn’t just write code; it writes code that passes the same review criteria we apply to human engineers,” Weller says. “Every branch, whether initiated by the CEO or a junior engineer, goes through the same automated quality gates and human code review.”

For business software provider Agiloft, “the biggest challenge isn’t technical, it’s organizational,” says Noe Ramos, vice president of AI operations. Agiloft is building an AI-native development capacity across every business function, embedding it into processes such as finance, human resources, and professional services.

“Most companies, including ours, are still learning where work actually happens versus where they think it happens,” Ramos says. “Before you can extend AI into a business function, you have to understand the real workflow, not the documented one. That discovery work is underestimated almost everywhere.”

The company also had to work through “the natural friction of trust and adoption,” Ramos says. “The human variable, though critically important to AI, is always the rate-limiting factor in this case, not the technology.”

On the governance side, Agiloft has had to address issues such as access controls, identity management, and data handling. “We built a formal approval process and a structured use-case lifecycle to manage this, so AI doesn’t get introduced into business teams in ways that create technical debt or compliance exposure,” Ramos says.

The primary challenge for healthcare technology provider iCore was creating confidence among non-technical users rather than technical barriers, says Thiago Soares, COO. The company has actively supported AI-assisted development outside its engineering team, including operations and client success functions that are building automated reporting tools, internal workflow templates, and client onboarding checklists that previously required developer time to produce.

“Staff unfamiliar with AI-assisted development needed structured onboarding before adoption felt natural,” Soares says. “We addressed that through peer-led sessions, where early adopters demonstrated practical use cases relevant to each team’s specific workflows.

Uneven confidence levels can slow progress, especially when people worry about getting it wrong, Daly says. “Creating a safe space to learn is important,” she says. “Creating small teams to work together with peer-to-peer support and encouraging shared learning has also been helpful to support progress through practical application, which is where we see people learn best.”

Governance is another key consideration, particularly in regulated industries with strict data handling protocols, such as healthcare. “Successful organizations set out explicit boundaries within which machine-generated code can be developed outside of formal development pathways,” Soares says.

IT leadership at iCore has built the guardrails needed to make expansion of vibe coding safe, “defining data access boundaries and compliance checkpoints that align with our commitment to trust and security,” Soares says. “That governance foundation is what allows cloud-driven innovation to move forward without creating the regulatory exposure healthcare environments cannot afford.”

The future of coding

Organizations that are moving vibe coding beyond software development teams are aiming to expand these efforts further.

At iCore, expansion of vibe coding into marketing and human resources functions is already under way, Soares says, focused on content workflows and documentation automation. These are “areas where vibe coding delivers efficiency without touching sensitive clinical or compliance infrastructure,” he says.

Agiloft plans to expand vibe coding further and make it into a standard practice “carefully and iteratively,” Ramos says, moving from isolated use cases toward a cohesive AI operating model with a shared infrastructure and AI-literate teams across functions, rather than just AI-enabled tools scattered across departments.

“That said, expansion for us means scaling what works, not scaling the toolset,” Ramos says. “We track every AI initiative through a structured lifecycle, from intake through decommission, precisely to avoid accumulating a stack of underutilized capabilities. Every function will eventually have embedded AI, but the goal is for those capabilities to be connected, governed, and [properly] used, not just deployed.”

EnFi is expanding from engineering-adjacent roles, such as product development and customer success, to other functions. “The same pattern — describe what you want, see it built, review and decide — applies to internal tooling, reporting, operational workflows, and documentation,” Weller says.

Organizations that master AI-assisted development internally “will be the ones capable of deploying AI-assisted workflows to their customers with the quality, governance, and reliability that regulated industries demand,” Weller says. “The internal practice is the proof point.”