Privacy Is The New Productivity Tax: How Data Fears Are Slowing Enterprise AI Adoption

New research shows data privacy ranks as Americans' second-highest artificial intelligence (AI) concern.

A new joint study by Cybernews and nexos.ai reveals that data privacy is the second-greatest concern for Americans regarding AI. This finding highlights a costly paradox for businesses: As companies invest more effort into protecting data, employees are increasingly likely to bypass security measures altogether.

The study analyzed five categories of concerns surrounding AI from January to October 2025. The findings revealed that the category of “data and privacy” recorded an average interest level of 26, placing it just one point below the leading category, “control and regulation.” Throughout this period, both categories displayed similar trends in public interest, with privacy concerns spiking dramatically in the second half of 2025.

Žilvinas Girėnas, head of product at nexos.ai, a secure all-in-one AI platform for enterprises, explains why privacy policies often backfire in practice.

“This is fundamentally an implementation problem. Companies create privacy policies based on worst-case scenarios rather than actual workflow needs. When the approved tools become too restrictive for daily work, employees don't stop using AI. They just switch to personal accounts and consumer tools that bypass all the security measures,” he says.

The enterprise privacy tax definition

The privacy tax is the hidden cost enterprises pay when overly restrictive privacy or security policies slow productivity to the point where employees circumvent official channels entirely, creating even greater risks than the policies were meant to prevent.

Unlike traditional definitions that focus on individual privacy losses or potential government levies on data collection, the enterprise privacy tax manifests as lost productivity, delayed innovation, and ironically, increased security exposure.

When companies implement AI policies designed around worst-case privacy scenarios rather than actual workflow needs, they create a three-part tax:

• The time tax. Hours get lost navigating approval processes for basic AI tools.

• The innovation tax. AI initiatives stall or never leave the pilot stage because governance is too slow or risk averse.

• The shadow tax. When policies are too restrictive, employees bypass them (e.g., using unauthorized AI), which can introduce real security exposure.

“For years, the playbook was to collect as much data as possible, treating it as a free asset. That mindset is now a significant liability. Every piece of data your systems collect carries a hidden privacy tax, a cost paid in eroding user trust, mounting compliance risks, and the growing threat of direct regulatory levies on data collection,” says Girėnas. 

“The only way to reduce this tax is to build smarter business models that minimize data intake from the start,” he continues. “Product leaders must now incorporate privacy risk into their ROI calculations and be transparent with users about the value exchange. If you can't justify why you need the data, you probably shouldn't be collecting it,” he adds.

How the privacy tax fuels a shadow AI crisis

The rise of shadow AI is mainly due to strict privacy rules. Instead of making things safer, these rules often create more risks. Research from Cybernews shows that  59% of employees admit to using unauthorized AI tools at work, and worryingly, 75% of those users have shared sensitive information with them.

“That’s data leakage through the back door,” says Girėnas. “Teams are uploading contract details, employee or customer data, and internal documents into chatbots like ChatGPT or Claude without corporate oversight. This kind of stealth sharing fuels invisible risk accumulation: Your IT and security teams have no visibility into what’s being shared, where it goes, or how it’s used.”

Meanwhile, concerns regarding AI continue to grow. According to a report by McKinsey, 88% of organizations claim to use AI, but many remain in pilot mode. Factors such as governance, data limitations, and talent shortages are impacting the ability to scale AI initiatives effectively.

“Strict privacy and security rules can hurt productivity and innovation. When these rules don't align with actual work processes, employees will find ways to get around them. This increases the use of shadow AI, which raises regulatory and compliance risks instead of lowering them,” says Girėnas.

Practical steps forward

To counter this cycle of restriction and risk, Girėnas offers four practical steps for leaders to transform their AI governance:

1. Provide a better alternative. Give the employees secure, enterprise-grade tools that match the convenience and power of consumer apps.

2. Focus on visibility, not restriction. Shift focus to gaining clear visibility into how AI is actually being used across the organization.

3. Implement tiered data policies. A “one-size-fits-all” lockdown is inefficient and counterproductive. Classify data into different tiers and apply security controls that match the sensitivity of the information.

4. Build trust through transparency. Clearly communicate to employees what the security policies are, why they exist, and how the company is working to provide them with safe, powerful tools.

ABOUT NEXOS.AI

nexos.ai is an all-in-one AI platform to drive secure, organization-wide AI adoption. Through a secure AI Workspace for employees and an AI Gateway for developers, nexos.ai enables companies to replace scattered AI tools with a unified interface that provides built-in guardrails, full visibility, and flexible access controls across all leading AI models — allowing teams to move fast while maintaining security and compliance. Headquartered in Vilnius, Lithuania, nexos.ai is backed by Evantic Capital, Index Ventures, Creandum, Dig Ventures, and a number of notable angels, including Olivier Pomel (CEO of Datadog), Sebastian Siemiatkowski (CEO of Klarna), through Flat Capital, Ilkka Paananen (CEO of Supercell), and Avishai Abrahami (CEO of Wix.com).