The uneven adoption of AI at work: Insights from the latest large-scale survey

AI adoption is steadily increasing. The survey reveals that 43.5% of professionals still do not use AI in their work, which is a significant improvement from 56% in last year’s pilot survey^[1] - marking a 12% increase in AI adoption over just one year.

Among those who use AI, the most common applications include:

• Content creation (24.1%)
• Data analysis and processing (22.6%)
• Assisting in creative processes (18.1%)
• Exploring and understanding new concepts (17.9%)
• Coding-related tasks (17.7%)

While these areas show strong momentum, other uses—like automating repetitive tasks (9.5%) and enhancing security measures (0.6%)—remain underutilized, signaling untapped potential. Meanwhile, a notable 5.4% of respondents indicated “other” tasks, highlighting niche and emerging use cases.

The survey shows that employees perceive their companies to be at different stages of AI integration:

• Only 26% have implemented comprehensive AI workplace policies.
• An additional 26% have policies, but they are insufficient.
• Shockingly, 23.7% have no policies at all, leaving employees without clear guidelines for AI usage.
   

Simultaneously, employees are voicing a strong need for upskilling:

• While 40.3% want more training on AI tools, only 9.8% feel they have received adequate upskilling.
   

This skills gap is concerning. Without targeted training, many professionals risk being left behind as AI continues to advance.

The analysis conducted by VTT reveals notable gender differences regarding concerns about AI-related bias and discrimination:

• Women are significantly more concerned:
  • 25.1% are somewhat concerned compared to 19.2% of men.
  • 4.5% of women report being extremely concerned, nearly double the 2.4% of men.
• In contrast, 67.2% of men report no concern, compared to 54.8% of women.

Interestingly, individuals who chose “other” or preferred not to disclose their gender reported higher uncertainty, with 55.6% selecting "somewhat concerned."

These findings emphasize the importance of addressing AI fairness and inclusivity to build trust across all demographic groups.

AI adoption has also sparked concerns about employee privacy, particularly in Finland, where strong traditions protect workers' personal communications. According to the survey:

• 42% of respondents are at least somewhat concerned about privacy when AI is implemented at their workplace.

In comparison, concerns about bias and discrimination are lower but still noteworthy, with 20.7% expressing some level of concern. These results highlight the need for companies to proactively address privacy protections and fairness concerns as they implement AI tools.

Our analysis of the TEK survey paints a clear picture of both progress and challenges in AI adoption at workplaces. To ensure no professional or company is left behind, the following actions are essential:

1. Develop tailored and comprehensive AI policies to provide clear guidance and foster ethical AI usage.
2. Prioritize upskilling employees so they can leverage AI’s full potential.
3. Proactively address privacy, bias, and inclusivity concerns to build trust and confidence in AI systems.

With AI adoption growing steadily—demonstrated by a 12% increase in just one year—the momentum is clear. However, the uneven take-off highlights the need for strategic action to harness AI’s transformative power for all workers and industries.

AI is no longer a future possibility; it is today’s reality. Companies that act now to close the gaps in policy, training, and fairness will not only boost productivity but also ensure sustainable and inclusive adoption of AI in the workplace.

Arash Hajikhani, VTT & Mikko Särelä, TEK

[1] Eight findings on how generative AI is perceived in the Finnish workforce, Hajikhani & Cole, 2023, https://www.vttresearch.com/en/news-and-ideas/eight-findings-how-generative-ai-perceived-finnish-workforce