The AI at Work Adoption Gap: 90% of Workers Use It, Most Still Don’t Trust It

Executive Summary

AI has moved beyond the realm of engineers and tech specialists. Today, it’s part of everyday work across industries, roles, and generations. In a Udacity survey of 2,000 workers, 9 in 10 reported using AI tools on the job, underscoring just how deeply integrated these technologies have already become.

Yet trust remains a major issue. Three out of four workers frequently abandon AI tools mid-task, citing concerns about accuracy, time spent refining outputs, and poor workflow fit. Nearly half also question the quality of AI-assisted work produced by their colleagues. The paradox is clear: workers are embracing AI, but they don’t fully believe in it.

At the same time, companies are lagging behind. Many don’t provide AI tools, clear policies, or meaningful training, leaving employees to experiment on their own, often paying out of pocket or resorting to unauthorized tools. This disconnect highlights a missed opportunity. While adoption is already high, organizations risk losing both productivity and trust if they fail to support workers with the resources and organizational culture needed for responsible and effective AI use.

Key Insights at a Glance

• Nearly ubiquitous use: 90% of workers say they have used some type of AI tool in their work.

• 3 in 4 workers regularly abandon AI tools mid-task, most commonly because their outputs lack accuracy or quality.

• Companies lag in support: Nearly half of workers say their employer doesn’t pay for any AI tools.

• Almost half (45%) say they do not trust their colleagues’ work when they know AI aided its production.

• Generational paradox: Gen Z feels the most comfortable with how AI may impact their careers, but they are also the most critical of others who use it.

• Gender gap: Women are more open to AI use but feel less supported and empowered to use it.

1. Nearly Everyone’s Using AI — But Trust is Low

The vast majority (90%) of respondents reported using AI tools in their work, with the most common applications being:

1. Writing emails, reports, and content (51%)

2. Research (51%)

3. Coding and technical tasks (50%)

4. Data analysis and insights (34%)

5. Productivity management (28%)

6. Creative work (25%)

Yet, trust is a major barrier:

• 3 out of 4 workers say it’s common for them to abandon AI mid-task, most often because the results lack accuracy or quality.

• 45% do not trust the quality of a colleague’s deliverable if they know it was created with the help of AI.

• 34% think less positively of colleagues who regularly use AI in their work.

• 36% would rather colleagues avoid AI use in their deliverables altogether.

“The data shows a clear gap between adoption and trust. With 9 in 10 workers using AI but 3 in 4 abandoning tasks due to poor outputs, the issue isn’t access—it’s execution. This points to a pressing need for skills in prompt engineering, context setting, and critically evaluating AI-generated content. Mastering how to refine and validate outputs quickly is the key to moving from casual experimentation to tangible productivity gains. Professionals who build fluency in these skills will be the ones to unlock the true potential of these tools.”Joseph Fontaine — AI Education Lead, Udacity

2. The Gen Z AI Paradox: Most Adaptable, Most Critical

Among all age groups, Gen Z feels the most comfortable with how AI may impact their careers:

• 59% feel adequately supported by their employer to adapt if AI significantly changes or replaces parts of their job (vs. 52% of Millennials, 45% of Gen X, and 40% of Boomers).

Simultaneously, they are also the most critical of AI use in the workplace:

• 45% would prefer colleagues not use AI in their work (vs. 34% of Millennials, 23% of Gen X, and 29% of Boomers).

• 42% think less positively of AI-using coworkers (vs. 33% of Millennials, 27% of Gen X, and 24% of Boomers).

This paradox suggests that Gen Z’s AI fluency may actually make them more aware of its limitations and the need for their coworkers to gain new skills to use the tools effectively.

“Gen Z’s stance is a reflection of their digital fluency. While 59% feel supported by their employers in adapting to AI, their skepticism towards colleagues’ AI-assisted work is notable, with 45% preferring it not be used at all by their colleagues. This isn’t a contradiction; it’s a higher standard. They recognize the difference between using AI and using it well. Their perspective highlights the growing importance of sophisticated AI literacy to ensure that AI-assisted work meets a high bar for quality and reliability.”Joseph Fontaine — AI Education Lead, Udacity

3. Companies Aren’t Keeping Pace

The workplace is in the middle of an AI surge (90% of workers say they use AI on the job), yet corporate support is far from keeping up. Many organizations are still operating as if AI is optional, leaving employees to bridge the gap themselves.

Nearly half of workers (47%) say they don’t receive sufficient resources and support to use AI effectively, and 42% report a lack of clear AI-use policies.

The gaps extend to basic access. Almost half of all workers (45%) say their company doesn’t pay for any AI tools. The numbers are even more striking among managers: 72% report having paid out of pocket for AI tools they need for work.

The combination of high demand and low institutional support has created a shadow IT phenomenon where nearly a third of workers (32%) use unauthorized tools. This presents significant security, compliance, and quality control risks that organizations need to address in the AI economy.

“The fact that 32% of workers are using unauthorized AI tools is a direct result of a leadership vacuum. This ‘shadow IT’ phenomenon isn’t just a security risk; it’s a clear signal that employees are desperate for tools their leaders haven’t provided. Effective AI integration requires active change management, starting at the top. Leaders must not only establish clear governance and provide secure, approved tools but also model responsible AI use themselves. This is about building a culture of innovation within a secure framework. Developing that strategy—one that balances empowerment with oversight—is a critical executive skill and the core focus of programs like our AI for Business Leaders Nanodegree program.”  Joseph Fontaine — AI Education Lead, Udacity

4. Women Trust AI More, But Are Hesitant to Admit Using It

Women stand out as more open and trusting toward AI use in the workplace compared to men. They are less likely to doubt the quality or accuracy of AI-assisted work from colleagues (37% vs. 47% of men), less likely to think negatively of coworkers who regularly use AI (26% vs. 37%), and less likely to prefer colleagues avoid AI use altogether (29% vs. 38%).

Yet, despite this greater openness, women are more hesitant about their own AI use. Just 55% say they feel comfortable admitting to a manager that they used AI to complete a task or deliverable, compared to 66% of men. This gap suggests that while women may be more trusting to the technology itself, they may be concerned about how their own use of AI is perceived, especially by those in positions of authority.

“The data reveals a cultural double standard that organizations must address. While women are more trusting of AI-assisted work from colleagues than men are, they are significantly less comfortable admitting their own use of it to management. This hesitation is often rooted in a justified fear that their work will be perceived as less authentic or less their own—a scrutiny not as commonly applied to men. To counter this, leaders must proactively reframe the narrative. They need to champion AI adoption as a sign of ingenuity and adaptability, not a shortcut. Strong, clear messaging that validates AI-assisted work as a powerful driver of productivity is essential to ensure that all employees, especially women, feel empowered to use these tools without penalty.”Joseph Fontaine — AI Education Lead, Udacity

5. Implications and Recommendations

For Organizations

1. Close the Tool Gap With 45% of companies not providing any paid AI tools, organizations must urgently evaluate and invest in appropriate AI solutions. The current situation where 47% of workers pay out of pocket is unsustainable and creates inequality.

2. Establish Clear Governance The fact that 32% use unauthorized tools indicates a desperate need for clear AI policies, approved tool lists, and secure alternatives to shadow IT solutions.

3. Invest in Quality Training With 74% abandoning AI mid-task and 26% struggling with prompting, comprehensive training programs are essential. 

Focus on:

• Effective prompting techniques

• Understanding AI limitations

• Quality assessment of AI outputs

• Integration with existing workflows

4. Address Cultural Resistance The trust deficit and negative perceptions of AI use require active cultural intervention. Leaders must model appropriate AI use and create psychologically safe environments for AI experimentation.

5. Bridge the Gender Gap Organizations must examine why women feel less supported despite being more open to AI, ensuring equal access to tools, training, and advancement opportunities in an AI-enhanced workplace.

For Workers

1. Develop AI Literacy Understanding both capabilities and limitations of AI tools is crucial for effective use and maintaining professional credibility.

2. Focus on Quality Control Given accuracy concerns, workers must develop strong skills in reviewing, validating, and enhancing AI outputs rather than accepting them wholesale.

3. Maintain Transparency 1 in 3 workers don’t feel comfortable admitting the use of AI in their work. Transparency about AI use can help build trust and establish best practices.

“There’s clearly a need to move beyond ad-hoc AI usage and build a truly AI-capable workforce. With 74% of workers abandoning AI tools mid-task, the demand for structured training is undeniable. A successful AI transformation requires a multi-layered approach that equips every level of the organization. It must start at the top, with our AI for Business Leaders program empowering executives to build a clear strategy and governance model. From there, you build universal buy-in and a common language with foundational training like our Agentic AI Fluency course for the entire organization. Finally, you empower your technical practitioners to build and deploy custom solutions with advanced programs like the Agentic AI Nanodegree program. This is how an organization moves from scattered use to a cohesive strategy that drives measurable productivity and innovation.” Joseph Fontaine — AI Education Lead, Udacity

6. Looking Forward: The Path to AI Maturity

The current state of AI in the workplace resembles the early days of internet adoption—widespread experimentation coupled with institutional uncertainty. Our research suggests we’re in a critical transition period where individual innovation outpaces organizational adaptation.

Success in this new era requires more than just deploying AI tools. It demands:

• Thoughtful integration strategies that address accuracy and workflow concerns

• Comprehensive training that goes beyond basic tool usage

• Cultural transformation that balances innovation with quality

• Equitable support systems that don’t leave any demographic behind

• Clear governance that provides security without stifling innovation

The organizations that successfully navigate these challenges won’t just adopt AI—they’ll create environments where human judgment and AI capabilities complement each other, where trust is maintained through transparency and quality, and where all workers feel empowered to leverage these tools effectively.

---

Udacity changes lives, businesses, and nations by creating job-ready digital skills at scale. With over a decade of experience, Udacity provides comprehensive solutions to address widespread digital talent shortages impacting growth, productivity, and innovation.