Follow ZDNET: Add us as your preferred source Google.
Important points of ZDNET
- Despite mass adoption of AI, employee trust in technology has declined.
- One factor could be lack of training.
- Businesses are looking for ways to reduce frustration.
For all the processes and workflows where AI saves time and increases efficiency, there are six where Tabby Farrar’s team feels technology doesn’t help.
Farrar is head of search at UK-based SEO and web design agency Candour. In digital marketing, like almost every other industry, there is a heated debate on the topic of AI. And while her team is eager to take advantage of the benefits of working faster, more efficiently, and reclaiming the time they were spending on less interesting tasks, it doesn’t always work that way.
Additionally, 95% of AI business applications fail. The reason is as follows
AI can generate for clients who do not have a product lifestyle image, but it will hallucinate or miss key points when creating a data summary. Refining the prompts for assigning categories to datasets could have been very time-consuming, and Farrar might have been better off doing the task manually.
“As a manager, I try to get my team more involved in AI stuff, because AI is the future of so many industries,” Farrar said. On the other hand, “there are so many people who say, ‘I wasted two hours of my day trying to make this work.'”
Farrar and her team are not alone in navigating the gap between the promise of AI and what it can actually do, and perhaps losing trust along the way.
Worker unrest causes real problems
A January survey by workforce solutions company ManpowerGroup found that employee confidence in AI fell for the first time in three years, dropping 18%, but adoption rates rose 13% year over year. A divergence in these numbers not only signals that the honeymoon period with AI is over, but also could serve as a wake-up call for organizations about how they implement AI tools in the workplace.
“You can’t have a workforce that’s threatened and you can’t be fully productive, and that insecurity is going to cause serious problems,” said Mara Stephan, vice president of global insights at ManpowerGroup.
Other studies paint a similar portrait of amputation. A November EY report found that while nine out of 10 employees are using AI in their work, only 28% of organizations are able to translate it into “high-value outcomes.”
“Our research shows why: Employees may be saving a few hours here and there, but nothing fundamentally changes the way they work or do business,” the report said.
For some, part-time work prevents this decline in worker confidence.
Related article: 5 ways you can stop testing AI and start scaling it responsibly in 2026
Randall Tinfow, CEO of Scranton, Pennsylvania-based AI-powered learning platform REACHUM, estimates that he spends about 20 hours out of a 70-hour work week vetting AI tools and partners to ensure they are not indiscriminately exposed to employees.
Platforms like Claude Code are saving software developers a lot of time with REACHUM, but not all are as effective. Tinfow believes there is a disconnect between how some AI tools are marketed and what they can actually do.
Even working at a company built around AI, Tinfow’s team encountered issues with certain AI tools not working for tasks like generating text in images.
“There’s so much noise that I don’t want the team to get distracted by it, so I’m the one who looks at something, decides whether it’s reasonable or garbage, and hands it to the team to work with,” Timphou said.
increase confidence
This discrepancy between expectations and reality may be a key reason for the decline in trust, said Kristin Ginn, founder of trnsfrmAItn, an organization that works with companies on AI adoption, focusing on the human workforce involved.
Marketing demos make everything look easy, but business leaders need to make sure their employees understand the trial and error and refinement that may lie ahead.
Psychological factors are also involved. A ManpowerGroup survey found that 89% of respondents were satisfied with their current role. For many people, work has been done in a unilateral way for a long time.
Related article: 5 ways rules and regulations can guide AI innovation
“If you start thinking about how to use AI for the same task, you suddenly have to expend more mental effort to figure out how to do this in a completely different way. That loss of routine and confidence in your way of doing things can also cause you to revert back to your human nature to avoid change,” Ginn said.
Stephen also discussed the role that proper training plays in maintaining confidence. More than half of respondents (56%) reported that they had not received any recent training or instruction (57%).
“Organizations and businesses that figure out how to deal with that, how to help their employees use technology, train, and feel better about the situation are going to benefit the most,” Stephan said.
looking for jewelry
Returning to digital marketing agency Khandua, Farah said the company has a variety of tactics to balance its pursuit of innovation with the day-to-day challenges of technology that still has a long way to go.
Candor set aside extra time to account for the fact that everyone is learning, framed experiments as “test and learn” to reduce stress, and appointed “champions” to keep abreast of AI developments. The agency’s chief marketing officer leads the training sessions, and Farah also conducts regular check-ins with the team. She also speaks openly about things that frustrate her at times.
Also: Prepare for these four scenarios to turn AI disruption into a career opportunity
Some initiatives have also proven fruitful, like the creation of Gemini Gem, trained on brand and tone of voice guidelines to create quotes that clients can tailor and approve for use in media. Candour’s innovation leaders use APIs from companies like OpenAI to build tools that more specifically meet enterprise needs. Farrar explained how people’s attitudes towards AI images quickly changed for the better after the launch of Google’s Nano Banana.
However, there is still a long way to go.
“If I’m going to use these tools and put some of my work aside, I want to trust that they’re going to do as good a job as I do,” Farrar said.
