Workers may feel confident in their ability to recognize AI-generated content, but that confidence is increasingly at odds with reality. A new national survey of more than 1,000 U.S. workers from AI resume builder Resume Now® finds that 74% say they can identify AI content, yet nearly half (48%) fail to do so when tested, revealing a widening gap between perception and reality.
As AI becomes more embedded in workplace communication, the line between human and machine-generated content is increasingly blurred. According to The AI Confidence vs. Reality Report, that uncertainty is already undermining worker confidence, with 65% saying that getting it wrong would make them less confident in their ability to identify AI in the future.
Key Insights:
Confidence vs. accuracy in AI detection: 74% of workers say they can identify AI-generated content, yet nearly half (48%) fail to do so when tested.
Misidentification is already widespread: 66% of workers say they have mistaken AI-generated content for human-written work.
AI content is a workplace norm: 49% of workers say they encounter AI-generated content at least weekly.
Workers suspect AI is involved in everyday communication: 42% of workers assume that even routine workplace messages, from emails to chat conversations, involve AI in some capacity.
Confidence is fragile when mistakes happen: 65% of workers say misidentifying AI-generated content would reduce their confidence in spotting it in the future.
Hidden AI use can damage trust between colleagues: 56% of workers say their trust in a coworker would decrease if they learned content was AI-generated but presented as human-written.
“Many workers feel confident in their ability to recognize AI-generated content, but confidence doesn’t always translate into accuracy,” says Keith Spencer, career expert at Resume Now. “As AI content becomes harder to distinguish, employees may question their own judgment, teams may doubt the credibility of others and organizations face a greater risk of inaccurate content reaching clients or the public. What seems like a small byproduct of AI use can quickly escalate into challenges around trust, accountability and brand reputation.”
Workers Overconfident They Can Spot AI
When presented with two nearly identical workplace messages and asked to identify which was written by AI, workers struggled to correctly identify the AI-generated content. Both blurbs used similar professional language, structure, and tone to reflect a typical workplace scenario.
52% correctly selected the AI-generated blurb
48% chose the human-written version
Despite this low accuracy, most workers felt confident they made the right choice:
21% were very confident
53% were somewhat confident
20% were not very confident
6% were not confident at all
AI Misidentification Is Becoming a Routine Experience at Work
Mistaking AI content for human-written work is not a one-off occurrence. For many workers, it is happening repeatedly as AI becomes more integrated into daily workplace communication.
66% of workers say they have mistaken AI-generated content for human-written work
24% say this has happened many times
42% say it has happened once or twice
34% say it has not happened
AI Content Is Now a Regular Part of the Workplace
Nearly half of workers (49%) say they encounter AI-generated content at least weekly.
22% say they encounter AI-generated content daily
27% say a few times a week
15% say a few times a month
20% say rarely
16% say never
Workers Increasingly Assume AI Is Involved in Work Messages
When reading everyday workplace communication, from emails to chat messages, many workers now assume that AI is involved to some extent, creating uncertainty around whether messages reflect a colleague’s own voice or AI-assisted output.
42% assume workplace messages from their coworkers involve AI in some capacity
29% assume they are written by a human with AI assistance
9% assume they are mostly AI-generated and lightly edited
4% assume they are mostly AI-generated with little or no human editing
58% believe messages received at work are written by a human
Difficulty Identifying AI-Generated Content Reduces Worker Confidence
65% of workers say that mistaking AI-generated content for human-written would reduce their confidence in spotting it in the future.
20% say it would significantly reduce their confidence
45% say it would somewhat reduce their confidence
Undisclosed AI Use Risks Trust Between Colleagues
When AI use isn’t clearly communicated, it can create rifts in workplace relationships.
56% of workers say their trust in a coworker would decrease if they learned content they believed was written by a colleague was actually AI-generated
23% say their trust would decrease significantly
33% say their trust would decrease slightly
35% say it would not affect their trust
9% say it would increase their trust
Together, these findings show that while workers feel confident identifying AI-generated content, many are unable to do so accurately, creating risks for trust, decision-making, and the integrity of workplace communication.
To view the full AI Confidence vs. Reality Report, please click here.
Methodology
The findings in this report are based on a survey of 1,006 employed U.S. adults, conducted by Resume Now using Pollfish in February 2026. Respondents answered questions about their ability to identify AI-generated content, confidence in detecting AI in workplace communication, frequency of encountering AI-generated content, and attitudes toward AI transparency and trust at work. Question formats included multiple-choice and single-select questions.
Demographic Breakdown
Survey respondents reflected a broad cross-section of the U.S. workforce. Participants were 50% male and 49% female, with 1% identifying as non-binary or preferring not to answer. Generationally, 21% identified as Gen Z, 26% as Millennials, 26% as Gen X, and 27% as Baby Boomers.
There are no comments
Please login to post comments