Employees hide valuable AI workflows and knowledge not due to governance, but primarily from lack of organizational trust.

Made with Rinto — analyse your own content free

The Problem: Concealed AI Usage

Many employees are choosing not to share their valuable AI workflow innovations, resulting in significant economic consequences.

Physician AI Template Example

A physician developed an 'astoundingly good' DoximityGPT template but withheld it from struggling colleagues.

AI Hiding Statistics

A KPMG/University of Melbourne study found 57% of employees admitted hiding AI use at work.

AI Hides Solutions

Unlike traditional knowledge hoarding of problems, AI introduces the suppression of solutions and valuable workflow innovations.

Productivity Gains from AI

AI's most valuable productivity gains often come from individual experimentation, easily refined and kept private.

Governance Misses the Point

Many organizations respond to 'shadow AI' with governance, but this targets the wrong problem; organizational trust is the true driver.

Reasons for Hiding AI Use

Employees hide AI use due to a desire for competitive edge, worry of more work, job insecurity, policy concerns, and social stigma.

AI Stress-Tests Culture

The underappreciated challenge is how AI stress-tests organizational culture, not just easing adoption anxiety.

Confusing Failure Types

Organizations often confuse blameworthy deviance with praiseworthy exploratory testing, punishing needed behavior.

Research Findings on AI Hiding

A survey of 604 U.S.-based employees revealed key predictors of AI knowledge hiding, particularly organizational trust.

Employees Withholding AI Knowledge

Nearly one in three survey respondents (30.3%) intentionally withheld AI-related knowledge or techniques.

Perceived Value of Sharing AI

Nearly four in five employees agreed that sharing AI knowledge would improve team tasks, solve problems, and raise productivity.

Organizational Trust Predicts Hiding

Organizational trust was the strongest predictor of AI knowledge hiding, even after accounting for other factors.

Trust and Approved Tools Synergy

Trust mattered more when employees had approved AI tools; trust creates willingness, approved tools create opportunity.

Formal Infrastructure Insufficiency

Neither an AI policy nor approved tools alone predicted hiding; formal AI infrastructure cannot solve the broader pattern.

Qualitative Insights on Distrust

Open-ended responses revealed employees hid knowledge to maintain advantage or avoid being targeted for job cuts.

Meta-analysis on Psychological Safety

A meta-analysis of 104 studies corroborates psychological safety reduces knowledge hiding, while mistreatment increases it.

Why Employees Keep Quiet

Employees make a rational calculation about the costs of making their AI workflows visible to the organization.

What Leaders Can Do

Leaders must credibly commit to a new deal where employee sharing strengthens their position, not undermines it.

Building Trust for AI Disclosure

A framework to encourage AI knowledge sharing involves providing reasons to contribute, easy paths, and skill development.

1Earn the Disclosure You Want→2Stop Taxing Efficiency Gains→3Reward Multiplier Behavior→4Legitimize AI Experimentation→5Treat Disclosure as Contribution

Earn the Disclosure You Want

Provide clear guidance on AI use and remove ambiguity, allowing employees to focus on work rather than perception.

Stop Taxing Efficiency Gains

Establish an explicit norm for how AI-saved time will be used, showing employees the upside beyond extraction.

Reward Multiplier Behavior

Reward reusable workflows, peer adoption, and quality improvements rather than just individual productivity.

Legitimize AI Experimentation

Sanction exploration time and name it to convert AI tinkering into a sanctioned category of work.

e.g.3M's 15% rule (Post-it Note), Google's 20% time (AdSense), and Anthropic's 'side quests' are examples.

Treat Disclosure as Contribution

Managers' immediate reaction to disclosure is a decisive trust signal; treat it as something worth understanding.

Warning About Tooling

Sanctioned enterprise AI tools log employee actions, amplifying the relationship between trust and hiding.

AI ROI Accrues to Employee

The ROI from AI is currently being kept by the employee because knowledge doesn't scale without visibility.

Key Question for Leaders

The critical question is whether employees trust the organization enough to share what they are learning with AI.

▸ 6 Expand

APEX

Organizational Trust and AI Knowledge Hiding

Employees hide valuable AI workflows and knowledge not due to governance, but primarily from lack of organizational trust.

Made with Rinto — analyse your own content free

▸ 8 Expand

SECT

The Problem: Concealed AI Usage

Many employees are choosing not to share their valuable AI workflow innovations, resulting in significant economic consequences.

EVID

Physician AI Template Example

A physician developed an 'astoundingly good' DoximityGPT template but withheld it from struggling colleagues.

DATA

AI Hiding Statistics

A KPMG/University of Melbourne study found 57% of employees admitted hiding AI use at work.

SUP

AI Hides Solutions

Unlike traditional knowledge hoarding of problems, AI introduces the suppression of solutions and valuable workflow innovations.

EVID

Productivity Gains from AI

AI's most valuable productivity gains often come from individual experimentation, easily refined and kept private.

OPP

Governance Misses the Point

Many organizations respond to 'shadow AI' with governance, but this targets the wrong problem; organizational trust is the true driver.

▸ 1 Expand

JUST

Reasons for Hiding AI Use

Employees hide AI use due to a desire for competitive edge, worry of more work, job insecurity, policy concerns, and social stigma.

DATA

Social Stigma of AI Use

An Anthropic study found 69% of professionals mentioned social stigma around AI use at work.

INSG

AI Stress-Tests Culture

The underappreciated challenge is how AI stress-tests organizational culture, not just easing adoption anxiety.

▸ 1 Expand

RISK

Confusing Failure Types

Organizations often confuse blameworthy deviance with praiseworthy exploratory testing, punishing needed behavior.

EVID

Stanford AI Adoption Challenges

A Stanford Digital Economy Lab study found 77% of AI adoption challenges were not technology-related, including earning trust.

▸ 7 Expand

SECT

Research Findings on AI Hiding

A survey of 604 U.S.-based employees revealed key predictors of AI knowledge hiding, particularly organizational trust.

DATA

Employees Withholding AI Knowledge

Nearly one in three survey respondents (30.3%) intentionally withheld AI-related knowledge or techniques.

DATA

Perceived Value of Sharing AI

Nearly four in five employees agreed that sharing AI knowledge would improve team tasks, solve problems, and raise productivity.

▸ 3 Expand

SUP

Organizational Trust Predicts Hiding

Organizational trust was the strongest predictor of AI knowledge hiding, even after accounting for other factors.

DATA

Trust Quartile Comparison

Employees in the lowest quartile of trust were nearly four times more likely to hide AI knowledge (47% vs. 14%).

DATA

Psychological Safety and Hiding

A similar pattern held for psychological safety (45% in low safety vs. 17% in high safety hid AI knowledge).

INSG

Trust Creates Psychological Safety

Trust likely reduces knowledge hiding by fostering psychological safety, allowing open sharing without fear.

INSG

Trust and Approved Tools Synergy

Trust mattered more when employees had approved AI tools; trust creates willingness, approved tools create opportunity.

OPP

Formal Infrastructure Insufficiency

Neither an AI policy nor approved tools alone predicted hiding; formal AI infrastructure cannot solve the broader pattern.

EVID

Qualitative Insights on Distrust

Open-ended responses revealed employees hid knowledge to maintain advantage or avoid being targeted for job cuts.

EVID

Meta-analysis on Psychological Safety

A meta-analysis of 104 studies corroborates psychological safety reduces knowledge hiding, while mistreatment increases it.

▸ 3 Expand

SECT

Why Employees Keep Quiet

Employees make a rational calculation about the costs of making their AI workflows visible to the organization.

RISK

Reputational Cost

Employees fear AI use makes them seem less capable or that their AI-generated work will be discredited.

RISK

Workload Cost

Efficiency gains from AI are often treated as spare capacity, leading to more work rather than better work.

▸ 1 Expand

RISK

Replaceability Cost

Employees fear their AI methods can be extracted, documented, and used to replace or automate their job.

DCSN

Prudent Career Strategy

Texas A&M professor Matthew Call advises employees to use personal AI tools to retain valuable knowledge.

▸ 2 Expand

SECT

What Leaders Can Do

Leaders must credibly commit to a new deal where employee sharing strengthens their position, not undermines it.

▸ 5 Expand

FRMW

Building Trust for AI Disclosure

A framework to encourage AI knowledge sharing involves providing reasons to contribute, easy paths, and skill development.

1Earn the Disclosure You Want→2Stop Taxing Efficiency Gains→3Reward Multiplier Behavior→4Legitimize AI Experimentation→5Treat Disclosure as Contribution

▸ 3 Expand

STEP

Earn the Disclosure You Want

Provide clear guidance on AI use and remove ambiguity, allowing employees to focus on work rather than perception.

TIP

Use Lightweight Sharing Methods

Convert private methods into reusable artifacts using lightweight templates, short demos, and 'show me how' sessions.

TIP

Build Disclosure Standing

Attach contributors' names to adopted workflows, ensuring disclosure builds a person's standing, not erases it.

EVID

Structured Conversations Impact

Structured conversations about techniques produced average sales gains exceeding 15% in a field experiment.

▸ 1 Expand

STEP

Stop Taxing Efficiency Gains

Establish an explicit norm for how AI-saved time will be used, showing employees the upside beyond extraction.

JUST

Why Taxing Gains is Harmful

If saved hours convert to more undesirable work, employees will rationally hide those efficiency gains.

▸ 3 Expand

STEP

Reward Multiplier Behavior

Reward reusable workflows, peer adoption, and quality improvements rather than just individual productivity.

JUST

Generic Leaderboards Amplify Hiding

Comparison-heavy climates can amplify retaliatory knowledge hiding, while learning climates reduce it.

EVID

Team-based Compensation Benefits

Team-based compensation generated substantially more knowledge transfer than individual commissions in a study of salespeople.

TIP

Give Specific Rewards

Rewards can include performance review credit, protected experimentation time, and a share of gains for wider workflow use.

STEP

Legitimize AI Experimentation

Sanction exploration time and name it to convert AI tinkering into a sanctioned category of work.

e.g.3M's 15% rule (Post-it Note), Google's 20% time (AdSense), and Anthropic's 'side quests' are examples.

▸ 2 Expand

STEP

Treat Disclosure as Contribution

Managers' immediate reaction to disclosure is a decisive trust signal; treat it as something worth understanding.

RISK

Disclosure as Unpaid Labor

Letting one person's disclosure become a standing obligation for training others converts honesty into unpaid labor.

TIP

Limit Sharing Cost

The discoverer demonstrates once; the company then owns documenting, distributing, and supporting the method.

▸ 1 Expand

RISK

Warning About Tooling

Sanctioned enterprise AI tools log employee actions, amplifying the relationship between trust and hiding.

JUST

Tooling and Trust Dynamic

If trust is low, approved tools are more likely to entrench hiding than to bring it into the open.

INSG

AI ROI Accrues to Employee

The ROI from AI is currently being kept by the employee because knowledge doesn't scale without visibility.

DCSN

Key Question for Leaders

The critical question is whether employees trust the organization enough to share what they are learning with AI.