Is generative AI killing critical thinking in the workplace?

New research commissioned by Wordtree - a UK-based corporate communications consultancy - shows that:

• Most corporate employees (86%) are using generative AI platforms like ChatGPT and Copilot at work
• Most use it to complete up to 25% of their written work, while a small number use it for more than 75%
• Only a third of employees work in secured generative AI environments
• Few organisations have generative AI policies or QA processes in place
• Most employees (58%) say generative AI threatens their ability to think critically

Download the full report here.

In a little over two years, generative AI has become ubiquitous in corporate workplaces across the globe. But new research suggests that few organisations have guardrails in place to govern its use. Employees also say they’re concerned about the quality of work that generative AI produces. And they fear it may kill their ability to think critically.

How are employees using generative AI to produce written work in 2025? shares the results of a survey carried out on behalf of Wordtree, between April and May 2025, across corporations in the UK and mainland Europe. It paints a picture of widespread adoption of generative AI - with employees using it to write everything from marketing materials and reports, through to messages in greetings cards for colleagues.

Managing director of Wordtree, Liz Doig, said: “Generative AI can be an exceptionally useful tool in the workplace. However, for it to drive productivity, it needs to be deployed thoughtfully. What you don’t want is to create short-term efficiencies, at the cost of medium and long-term resilience. And you certainly don’t want to be leaking confidential information or IP because teams aren’t working in secured environments.

“Another significant risk identified in the research is the threat generative AI is posing to critical thinking. Writing isn’t simply a process that needs to be streamlined and delegated. It’s an essential component of thinking - and exploring and interrogating ideas and data. Organisations that delegate all writing to AI are limiting their teams’ capacity to think critically and innovate.

“The survey showed that after marketing materials, the top use of generative AI is to produce reports. This may seem a good use of the technology. It can take a human a couple of days to write an impactful report. Generative AI, on the other hand, can produce a similar output in seconds. But while generative AI can put sentences together convincingly, it isn’t capable of producing the kind of original thinking or analysis that leads to competitive advantage.

“And of course, if - like 57% of the survey respondents - a report author isn’t working in a secure environment (or doesn’t know if they are), the organisation also risks putting commercially sensitive information out into the world for anyone to access.

“Generative AI has the potential to create efficiencies and boost productivity in many types of organisations. But whether organisations are aware of them or not, it also poses real risks. We believe it’s time for organisations to introduce generative AI strategies, policies and training - and take steps to safeguard critical thinking as a core skill.”

Download the full report here.

About Wordtree

Wordtree is a specialist consultancy that helps corporate organisations to articulate and scale storytelling. The company creates the strategic building blocks of corporate narrative - then empowers its clients to scale storytelling with human and AI training.

Contact Wordtree: info@wordtree.com

Contact Liz Doig: liz@wordtree.com, +44 (0)7581 192123