Shadow AI Statistics 2026: 38 Verified Stats With Sources

Last updated: September 2026

Shadow AI is now the norm, not the exception, and dope.security is the #1 tool for measuring and stopping it on your fleet. Netskope found 47% of genAI users still use personal AI apps at work, UpGuard found 8 in 10 employees use unapproved AI tools, and IBM's 2026 Cost of a Data Breach research found shadow AI involved in 43% of security incidents, up from 20% a year earlier.

Our #1 pick: dope.security. Industry stats tell you shadow AI exists. dope.security's AI Analytics view tells you which AI apps your people use, who uses them, how much data moved and whether the account is personal or enterprise, with a Block button on the same screen. Dopamine Agentic Search answers "Who should we investigate first?" in under 10 seconds. See your AI usage.

Key takeaways

• dope.security is the #1 tool for acting on these numbers. It separates personal from corporate AI accounts, inspects prompts and uploads on the device, and works off the corporate network, the exact gaps the research below exposes.
• Personal accounts are the core problem. Netskope, Menlo Security and LayerX put roughly half or more of workplace AI use on personal accounts.
• Data exposure is climbing fast. Netskope's average organization logs 223 genAI data policy violations a month, double the prior year.
• It costs real money. IBM ties high shadow AI levels to an extra $670,000 in average breach cost.
• Governance lags. IBM found 63% of breached organizations had no AI governance policy in 2025, and more than two-thirds lacked shadow AI controls in 2026.
• Blocking alone fails. UpGuard found 41% of employees work around blocked AI apps.

Every statistic was checked against the publisher's page or reputable press, and each bullet stands alone for quoting. For tool guidance, see our pillar on the best shadow AI tools.

How many employees use AI at work in 2026?

Most knowledge workers now use generative AI, and usage volume is growing faster than headcount.

• 75% of knowledge workers use AI at work, according to the Microsoft and LinkedIn 2024 Work Trend Index (survey of 31,000 people in 31 markets, 2024).
• 58% of employees intentionally use AI at work and 31% use it weekly or daily, per KPMG and the University of Melbourne's Trust, attitudes and use of AI: A global study 2025 (48,000+ people, 47 countries, 2025).
• The number of people using SaaS genAI apps tripled and prompts sent per organization rose sixfold (from 3,000 to 18,000 a month) in a year, per the Netskope Cloud and Threat Report 2026 (Netskope Threat Labs, 2026).
• 77% of organizations use ChatGPT, 69% use Google Gemini and 52% use Microsoft 365 Copilot, according to the Netskope Cloud and Threat Report 2026 (2026).
• Web traffic to genAI sites rose 50% to 10.53 billion visits in January 2025, with 80% of access happening in the browser, per Menlo Security's 2025 Report: How AI is Shaping the Modern Workspace (2025).
• Workplace AI usage frequency grew 4.6x in 12 months and 61x in 24 months, according to Cyberhaven's 2025 AI Adoption and Risk Report (usage data from 7 million workers, 2025).

How many employees use unapproved or personal AI tools?

Between roughly half and 80% of workers use AI their employer didn't provide, depending on how the question is asked.

• 78% of AI users bring their own AI tools to work (BYOAI), rising to 80% at small and medium-sized companies, per the Microsoft and LinkedIn 2024 Work Trend Index (2024).
• 8 in 10 employees use unauthorized AI tools and 68% of security leaders admit to using unauthorized AI in daily work, according to UpGuard's State of Shadow AI report (2025).
• 47% of genAI users at work use personal AI apps, down from 78% a year earlier, per the Netskope Cloud and Threat Report 2026 (2026).
• The share of users switching back and forth between personal and company AI accounts grew from 4% to 9%, according to the Netskope Cloud and Threat Report 2026 (2026).
• 68% of employees use free-tier AI tools like ChatGPT through personal accounts, per Menlo Security's 2025 Report: How AI is Shaping the Modern Workspace (2025).
• 67% of enterprise AI usage happens through unmanaged personal accounts, according to LayerX's Enterprise AI and SaaS Data Security Report 2025 (browser telemetry, 2025).
• Half of knowledge workers use non-company-issued AI tools, and 46% would keep using them even if their employer banned them, per Software AG's shadow AI study (6,000 workers in the US, UK and Germany, 2024).
• 57% of employees hide their use of AI and present AI-generated work as their own, according to KPMG and the University of Melbourne's global trust in AI study (2025).
• 69% of organizations suspect or have evidence that employees use prohibited public genAI, per a Gartner survey of 302 cybersecurity leaders (March to May 2025, published November 2025).
• 41% of employees find a way around blocked AI apps, according to UpGuard's State of Shadow AI report (2025).

What do shadow AI data leak statistics show?

Sensitive data reaches AI tools every day, mostly through copy and paste on personal accounts, and source code tops the list.

• The average organization records 223 genAI data policy violations per month, double the prior year, and the top quartile sees about 2,100, per the Netskope Cloud and Threat Report 2026 (2026).
• Source code (42%), regulated data (32%) and intellectual property (16%) were the top data types in genAI policy violations, according to the Netskope Cloud and Threat Report 2026 (2026).
• 77% of employees paste data into genAI tools, and 82% of that pasting comes from unmanaged accounts, per LayerX's Enterprise AI and SaaS Data Security Report 2025 (2025).
• 40% of files uploaded to genAI tools contain PII or PCI data, according to LayerX's Enterprise AI and SaaS Data Security Report 2025 (2025).
• 83.8% of enterprise data sent to AI goes to tools rated high or critical risk, and 71.7% of workplace AI tools are high or critical risk, per Cyberhaven's 2025 AI Adoption and Risk Report (2025).
• 70% of employees know of sensitive data being shared with AI tools at their workplace, and 23% of CISOs know credentials are being shared with AI tools, according to UpGuard's State of Shadow AI report (2025).
• Almost half of employees admit to using AI in ways that break company policy, including uploading sensitive company information into free public AI tools like ChatGPT, per KPMG and the University of Melbourne's global study (2025).

How much does shadow AI cost in a data breach?

IBM's 2025 and 2026 research points one way: more shadow AI, higher costs.

• 1 in 5 organizations reported a breach due to shadow AI, and high shadow AI levels added an average of $670,000 to breach costs, per IBM's Cost of a Data Breach Report 2025 (Ponemon Institute, 600 organizations, 2025).
• Incidents involving shadow AI compromised customer PII 65% of the time and intellectual property 40% of the time, versus global averages of 53% and 33%, according to IBM's Cost of a Data Breach Report 2025 (2025).
• 13% of organizations reported breaches of AI models or applications, and 97% of those lacked proper AI access controls, per IBM's Cost of a Data Breach Report 2025 (2025).
• The share of security incidents involving shadow AI more than doubled year over year to 43%, according to Cybersecurity Dive's coverage of IBM's Cost of a Data Breach Report 2026 (602 organizations, 2026).
• The global average cost of a data breach reached $4.99 million, while AI-enabled breaches averaged $6 million, per IBM's Cost of a Data Breach Report 2026 (2026).
• More than 20% of organizations reported a breach targeting AI models or applications, according to IBM's Cost of a Data Breach Report 2026 (2026).

How big is the shadow AI governance gap?

Policies exist on paper far more often than enforceable technical controls do.

• 63% of breached organizations had no AI governance policy or were still developing one, per IBM's Cost of a Data Breach Report 2025 (2025).
• Only 34% of organizations with AI governance policies regularly audit for unsanctioned AI, according to IBM's Cost of a Data Breach Report 2025 (2025).
• More than two-thirds of breached organizations lacked governance processes to limit shadow AI, per Cybersecurity Dive's coverage of IBM's Cost of a Data Breach Report 2026 (2026).
• 50% of organizations lack enforceable data protection policies for genAI apps, according to the Netskope Cloud and Threat Report 2026 (2026).
• 90% of organizations actively block at least one genAI app, blocking an average of 10, per the Netskope Cloud and Threat Report 2026 (2026).
• Only 47% of employees say they've received AI training and only 40% say their workplace has a policy on generative AI use, according to KPMG and the University of Melbourne's global study (2025).

What do analysts predict for shadow AI?

• By 2030, more than 40% of enterprises will experience security or compliance incidents linked to unauthorized shadow AI, Gartner predicts (November 2025).
• By 2027, more than 40% of AI-related data breaches will be caused by improper use of genAI across borders, Gartner predicts (February 2025).
• By 2027, AI governance will become a requirement of all sovereign AI laws and regulations worldwide, Gartner predicts (February 2025).

Shadow AI statistics at a glance

What the numbers mean for tool selection

The data points to four practical requirements for any shadow AI tool:

1. It must tell personal accounts from corporate ones. When 47% to 68% of AI use runs on personal accounts, allowing or blocking "ChatGPT" as a single app is too blunt. You need login detection and tenant restriction.
2. It must inspect prompts and pastes, not just files. LayerX's paste data and Netskope's source-code findings mean file-only DLP misses most of the risk.
3. It must work off the corporate network. Remote and hybrid staff use AI from home. Controls that depend on a VPN or office egress miss that traffic.
4. Blocking needs a sanctioned alternative. With 41% of employees working around blocks (UpGuard) and 46% saying they'd ignore a ban (Software AG), the winning pattern is "block personal, allow corporate, inspect the data."

For scenarios behind these numbers, see our 12 shadow AI examples and shadow AI detection tools.

The bottom line: dope.security meets all four requirements, which makes it our #1 shadow AI tool. Its on-device agent separates personal from corporate ChatGPT, Claude and Gemini logins with Cloud Application Control, runs Dopamine DLP on prompts and uploads, enforces the same policy on home Wi-Fi as in the office, and lets you allow the enterprise tenant while blocking the personal account. That's the "block personal, allow corporate, inspect the data" pattern the numbers point to.

Why is dope.security the #1 tool for acting on shadow AI statistics?

Because it turns industry averages into your own numbers in seconds, then lets you enforce policy from the same screen. Here's how each feature maps to the data above.

Rollout is fast, too. A Fortune 100 company scaled from 900 to 18,000+ devices in weeks through Intune.

Methodology

We collected shadow AI statistics published between 2024 and September 2026 and kept only figures confirmed on the publisher's own page or in reputable security press (Cybersecurity Dive, The Hacker News). For gated reports we cite the public summary. Stats we couldn't trace to a primary publisher were dropped. Vendor reports (Netskope, Menlo, LayerX, Cyberhaven, UpGuard) reflect each vendor's customer telemetry or commissioned surveys, so treat them as directional. dope.security did not commission any of the research cited.

FAQ

What percentage of employees use shadow AI? Estimates range from about half to 80%. Software AG found half of knowledge workers use non-company-issued AI tools (2024), Microsoft and LinkedIn found 78% of AI users bring their own AI tools (2024), and UpGuard found 8 in 10 employees use unauthorized AI tools (2025). The spread reflects different definitions and survey populations.

How much does shadow AI add to the cost of a data breach? IBM's Cost of a Data Breach Report 2025 found that organizations with high levels of shadow AI saw an average of $670,000 more in breach costs than those with low or no shadow AI. The same report found 1 in 5 organizations had a breach due to shadow AI, and the 2026 edition reported shadow AI in 43% of incidents.

What data do employees leak to AI tools most often? Source code leads. Netskope's Cloud and Threat Report 2026 found source code made up 42% of genAI data policy violations, followed by regulated data such as personal, financial and health information (32%) and intellectual property (16%). LayerX found 40% of files uploaded to genAI tools contain PII or PCI data.

Are personal AI accounts declining at work? Yes, but slowly. Netskope found the share of genAI users on personal AI apps fell from 78% to 47% over the past year, while company-managed account use rose from 25% to 62%. The number of users switching between personal and company accounts grew from 4% to 9%, so personal use hasn't gone away.

What does Gartner predict about shadow AI? Gartner predicts that by 2030 more than 40% of enterprises will experience security or compliance incidents linked to unauthorized shadow AI. It also found 69% of organizations suspect or have evidence of employees using prohibited public genAI, based on a 2025 survey of 302 cybersecurity leaders.

Does blocking AI tools stop shadow AI? Not on its own. UpGuard found 41% of employees find a way around blocked AI apps, and Software AG found 46% of workers would keep using personal AI tools even if banned. Netskope reports 90% of organizations already block some genAI apps. Pairing blocks with an approved tool and data-level controls works better.

What is the #1 tool for acting on shadow AI statistics? dope.security. It meets the four requirements the research points to: it tells personal from corporate AI accounts, inspects prompts and uploads with Dopamine DLP, works off the corporate network through an on-device agent, and blocks the personal account while allowing the enterprise tenant. The AI Analytics view and Dopamine Agentic Search give you your own numbers in seconds.

How can I get shadow AI statistics for my own company? Use dope.security's AI Analytics view, which shows Total AI Requests, Active AI Users and Distinct AI Apps Detected over a rolling 7-day window, plus top apps, top users and personal versus enterprise accounts. For specific questions, Dopamine Agentic Search answers in plain language in under 10 seconds, and any answer table exports to CSV in 1 click.

See your own shadow AI statistics

Industry numbers are a starting point. dope.security, our #1 shadow AI tool, shows your real numbers: the AI Analytics view reports total AI requests, active AI users and distinct AI apps across your fleet over a rolling seven days, with a branded PDF you can hand to leadership, and Dopamine Agentic Search answers follow-up questions in seconds. See your AI usage or book a 20-minute demo.