Early AI Adopters Are Now Becoming More Cautious About the Technology

Artificial intelligence was initially embraced by many professionals as a powerful way to save time, automate repetitive work and improve productivity. But as the technology becomes more deeply integrated into workplaces, some of its earliest users are beginning to question how much of their daily work should actually depend on AI.

The change does not necessarily mean that professionals have turned against artificial intelligence. Instead, early experience with the technology appears to have made some users more aware of its limitations, security concerns and potential impact on their careers.

Research from Thomson Reuters shows that AI is now firmly embedded in professional work. In its 2026 Future of Professionals report, 74% of professionals surveyed said they use AI several times a week. At the same time, 48% said they were concerned about AI negatively affecting the development of independent professional judgment.

Privacy and Security Are Becoming Major Concerns

For some of the earliest users, concerns are no longer theoretical.

Recent reports indicate that users of personal AI agents have started deleting or restricting these systems after experiencing privacy problems and unexpected behavior. Some users became uncomfortable after AI agents were given access to sensitive information, including email accounts and text messages.

These experiences highlight an important problem: the more powerful an AI assistant becomes, the more information and permissions it may require.

Giving an AI system access to calendars, emails, files and online accounts can make it considerably more useful. However, it can also increase the consequences of an error or security failure.

Professionals Are Asking Different Questions

Early adopters were initially focused on what AI could accomplish. Now, many professionals are also asking whether every task should be automated.

This is particularly important in professions where decisions require experience, responsibility and independent judgment.

Thomson Reuters found that professionals are dealing with a significant gap between AI adoption and organizational guidance. Around 40% of surveyed professionals reported receiving contradictory instructions about AI use from clients and company leadership.

That uncertainty can make employees reluctant to rely heavily on AI, particularly when they do not know who will be responsible if an AI-generated result turns out to be wrong.

The Early Adopters Once Saw Major Benefits

The shift in attitude is particularly interesting because early adopters were among the groups most enthusiastic about the technology.

Earlier Thomson Reuters research found that 70% of professionals said their teams were already seeing benefits from AI-powered tools, particularly for research, drafting, document summaries and automation. Among organizations considered AI leaders, that figure reached 94%.

This suggests that the current caution is not necessarily a rejection of AI.

Instead, professionals who have spent more time using the technology may simply have a clearer understanding of where it works well and where human expertise remains essential.

AI May Be Used More Selectively

Rather than abandoning artificial intelligence completely, some professionals may begin using it for specific tasks while keeping humans responsible for important decisions.

AI can be useful for summarizing large documents, organizing information, generating initial drafts and handling repetitive administrative work.

However, tasks involving legal responsibility, financial decisions, medical judgments or sensitive personal information require considerably more caution.

Thomson Reuters argues that general-purpose AI is not sufficient for high-stakes professional work without authoritative information, strong privacy protections, transparency and human oversight.

Fear of Job Loss Also Remains

Another reason for growing caution is employment.

A 2026 Randstad survey found that 80% of workers believe AI will affect their daily tasks. Almost half of respondents were concerned that the technology could benefit companies more than employees, while younger workers expressed particularly high levels of concern.

For professionals who adopted AI early, this creates a complicated situation. The same technology that makes their work faster could eventually change which skills employers value and how many people are required to perform certain tasks.

The Future May Be