A Tuesday (Nov. 10) blog post from cybersecurity firm Claroc reported that over 35,000 business applications built using Grok’s AI were accessible without authentication, including tools containing patient information, employee passwords and proprietary corporate records.
The discovery underscores a potential structural risk in the emerging market for AI-generated business software.
Claroc identified internal information belonging to 30 organizations beyond Grok developer SpaceXSI and X, including six Fortune 500 companies. Exposed materials allegedly included a sales territory map covering 594 accounts, an intensive care patient’s medical information and a timesheet application containing plaintext passwords.
The research does not establish that malicious actors exploited the exposed information. But it illustrates how an ordinary publishing decision can turn a productivity tool into a security exposure.
Read more: The CFO’s Guide to LLMjacking: When Stolen AI Becomes a 7-Figure Cloud Bill
AI Is Rewriting the Economics of Shadow IT
Traditional shadow IT involved employees adopting software outside approved procurement channels. AI app builders introduce a more consequential variation: Employees can now create the software themselves. A sales manager can build a customer dashboard. A healthcare administrator can generate a scheduling application. An operations team can create an internal database without commissioning a development project.
The PYMNTS Intelligence report “How Agentic AI Went From Zero to CFO Test Runs in 90 Days” found in September 2025 that nearly 7% of enterprise CFOs in the U.S. had deployed agentic AI in live finance workflows, while an additional 5% were running pilots.
Applications that previously required developers, security reviews and deployment approvals can now become operational through conversational interfaces. The productivity opportunity is substantial. So is the change in accountability.
In effect, AI can shift software expenses from development budgets into ongoing operational risk.
The PYMNTS Intelligence report “Vendors and Vulnerabilities: The Cyberattack Squeeze on Mid-Market Firms” found that hackers increasingly target middle market firms. These companies depend on third-party cloud providers, software-as-a-service platforms and managed service providers, which can leave them exposed.
Read also: Vibe Coding Breaks Into Banking Before Regulators Can React
IBM said over the summer that a higher percentage of companies were planning to up their security spending after finding out about the cyber capabilities of frontier AI models, and 68% of financial institutions increased their cyber budgets year over year, according to the “2025 State of Fraud and Financial Crime in the United States,” a PYMNTS Intelligence report produced in collaboration with Block.
The Grok findings concern one platform and do not establish that all AI-built applications share its vulnerabilities. But they reveal an important limitation in measuring AI productivity solely through applications created or development hours saved.
Grok representatives did not immediately reply to PYMNTS’ request for comment.
For enterprise leaders, the larger question is whether governance can become embedded in the development process rather than imposed afterward. AI is making software cheaper to build. Whether it makes businesses cheaper to run will depend on how much of that savings survives the cost of securing what employees create.
For all PYMNTS AI and digital transformation coverage, subscribe to the daily AI and Digital Transformation Newsletters.
