AI Saves Employees 2.3 Hours Daily, but 83% Fear Consequences of AI Errors

Artificial intelligence is helping employees save an average of 2.3 hours each day, but concerns about responsibility for AI-related mistakes remain high. A GoTo study conducted with Workplace Intelligence found that 83% of employees worry they could be blamed or even lose their jobs because of an AI error.

These concerns are also influencing how employees use the technology. Around 31% said they had delayed using AI because they were concerned about possible consequences if something went wrong. AI policies are less common among smaller organisations, with 24% of small businesses and 36% of midsize companies reporting that they have such policies, compared with 53% of enterprises.

The study was conducted between November 2025 and January 2026 and included 2,500 respondents from the US, Canada, UK, Ireland, Germany, Austria, Switzerland, India, Mexico and Brazil. The respondents consisted of 1,250 full-time knowledge workers and 1,250 IT decision-makers.

Businesses are already reporting some effects from AI mistakes. One in four IT leaders said AI errors had negatively affected customers or clients, while 23% reported financial damage linked to AI mistakes. Among organisations that had not experienced such an impact, 91% of IT leaders remained concerned that an AI error could affect their company in the future.

The study also found that some employees use AI to shift responsibility for mistakes. About 26% said AI can help them avoid accountability, while 25% said they consider it safer to blame AI than acknowledge their own error. Another 17% said they had already blamed AI for a mistake they had made.

Responsible use remains another concern. While 84% of respondents believe their companies could do more to promote responsible AI use, 55% said they do not always check AI-generated outputs, including when handling high-stakes work. In addition, 43% said they had used AI-generated material that they considered poor quality or believed could contain errors or fabricated information.

Workplace pressure is also shaping AI adoption. Six in ten respondents said they feel pressured to use AI to increase productivity, while 47% said AI usage is considered in performance reviews. About 42% would rather accept the possibility of an AI mistake than risk appearing slow or replaceable.

The study found that 31% of employees experience unspoken pressure to trust AI and remain quiet about mistakes. Around 14% said they had reported an AI error but were subsequently asked to stay silent about it.

The difference in preparedness is also evident in the adoption of agentic AI. While 46% of small businesses are using or testing such systems, the figure rises to 70% among enterprises. At the same time, 38% of SMB IT leaders said they were not prepared to manage these systems securely, compared with 14% of enterprise IT leaders.

As more routine work becomes automated, employees continue to identify human capabilities as important. According to the study, 52% selected creative thinking and adaptability as the most important human skills to retain.