# Workslop: The Hidden Tax of AI at Work > 40% of desk workers got 'workslop' last month: polished AI output with no substance. Here's what it costs Indian teams, and the skills that stop it. - Source: https://bodhih.com/workslop-ai-generated-work-india-teams-training/ - Publisher: Bodhih Training (Bengaluru, India; since 2008) - Last updated: 2026-10-08 - Contact: solutions@bodhih.com · +91 99000 11601 --- 40% of desk workers got 'workslop' last month: polished AI output with no substance. Here's what it costs Indian teams, and the skills that stop it. 40% of desk workers got 'workslop' last month: polished AI output with no substance. Here's what it costs Indian teams, and the skills that stop it. Key takeaways - Researchers at BetterUp Labs and the Stanford Social Media Lab coined "workslop" in 2025 for AI-generated work that looks polished but lacks the substance to move a task forward. - In their survey of U.S. desk workers, 40% said they had received workslop in the previous month, and respondents estimated that about 15% of the work they receive fits the description. - The same research estimated roughly two hours of rework per incident, or about $186 per employee per month. That works out to around $9 million a year for a 10,000-person company. These are self-reported estimates, so treat them as directional. - The damage is not only time. About half of recipients said they saw the sender as less creative, capable and reliable, and 42% saw them as less trustworthy. - In India, the gap behind the problem is measurable: a January 2026 Genius HRTech and Digipoll survey found 61% of professionals said their organisation had not given adequate guidance on using AI, and only 37% had received proper training. - Workslop is a skills problem, not a tool problem. The fix is training in AI literacy, critical thinking and writing, plus managers who can set a quality bar. Last updated 08 Oct 2026 · Bodhih Insights team The deck looked perfect. Twelve slides, crisp headings, a confident executive summary, even a tidy risk matrix. Ananya, a delivery head at a Pune GCC, opened it ten minutes before a client call and felt relieved. Then she read slide four. The "market sizing" cited a figure that appeared nowhere in the source report. Slide seven recommended a vendor the company had blacklisted two years ago. The summary contradicted the data underneath it. She spent the next ninety minutes quietly rebuilding the deck herself, and said nothing to the junior analyst who had sent it. Nobody logged the lost time. Nobody will. But Ananya had just paid a tax that now has a name: **workslop**. ## The Productivity Gain That Moves, Not Disappears Most conversations about AI at work focus on speed. A report that took three hours now takes twenty minutes. That is true, and it is why adoption has been so fast. A Salesforce India survey of 1,029 desk workers found that 94% were keen to build AI skills, and 61% were already using AI in their roles. Workslop is what happens when the speed is real but the quality check is missing. The sender saves two hours. The receiver loses two hours. The total effort in the organisation has not fallen; it has simply been **moved from the person who made the work to the person who has to trust it**. And because the receiver is usually more senior, the hours moved are also the more expensive ones. This is why many leaders report a strange feeling: everyone says they are more productive, but projects do not seem to close faster. The savings are being spent downstream, in review cycles, clarification calls and quiet rewrites that never appear on any dashboard. ## Why It Spreads Faster in Some Teams Than Others Workslop does not come from careless people. It comes from three very ordinary conditions. **The training gap.** The Genius HRTech and Digipoll data tells the story: most Indian professionals are using AI without much guidance. Salesforce's survey found that 40% of desk workers had spent less than five hours in total learning how to use it. A tool that fluent can produce confident-looking text from a vague prompt. Without training, there is no one to say what "good" looks like. **The ownership gap.** When someone writes a document themselves, they own every sentence. When AI writes it, ownership gets blurry. "The AI said so" quietly becomes an acceptable answer, and the sender forwards a draft they have never really read. **The manager gap.** Many managers have not set any rule about AI output at all. Is it fine to send an unedited draft? Must sources be checked? Should AI use be disclosed? Without an explicit standard, the standard becomes whatever the fastest person in the team does. ## The Cost That Doesn't Show Up on a Timesheet The most interesting finding in the research is not about hours. It is about trust. When a colleague sends you workslop, you do not just redo the work. You revise your opinion of them. About half of the people surveyed said the sender now seemed less creative, capable and reliable. Forty-two percent saw them as less trustworthy. Nearly a third said they would be less willing to work with that person again. For a team, that is the real bill. Collaboration runs on a basic assumption: *what you send me has been thought about.* Every workslop incident chips at that assumption. Over a year, a team where it happens often learns to double-check everything from everyone, and a team that double-checks everything has just rebuilt the slow, bureaucratic process that AI was supposed to remove. ## What Actually Prevents Workslop The answer is not banning AI, and it is not another policy PDF. The teams that avoid workslop tend to build three capabilities. **Everyone learns to use AI with judgment, not just speed.** The skill is not writing a clever prompt. It is knowing what to ask, recognising when an answer is plausible but wrong, and verifying before sending. This is the core of [AI literacy](https://bodhih.com/ai-literacy-training/), and it is what programs like Bodhih's [AI Literacy Training](https://bodhih.com/ai-literacy-training/) are designed to build, with hands-on practice on the team's own real work rather than generic demos. **Writing and thinking skills get stronger, not weaker.** AI amplifies whatever thinking it is given. A person who can structure an argument, state a point in two sentences and spot a gap in logic will get far better output, and will catch bad output faster. That is why [Business Writing with AI](https://bodhih.com/business-writing-with-ai-training/) and [Critical Thinking Training](https://bodhih.com/critical-thinking-training/) matter more now, not less. They are the quality filter that sits between the tool and the inbox. **Managers set and model a quality bar.** A team needs a simple, shared answer to "what must a person do before sending AI-assisted work?" Leaders who have been through something like [AI for Leaders](https://bodhih.com/ai-for-leaders-training/) can define that bar, ask better review questions, and respond to a weak draft with coaching instead of a silent rewrite, which is exactly what Ananya did not do. ## A Five-Line Standard Any Team Can Adopt This Week You do not need a program to start. Agree on these five rules as a team and write them somewhere visible: 1. **Own it.** If your name is on it, you have read every line. 2. **Check the facts.** Every number, name and citation gets verified against a source. 3. **Say what you want.** State the decision or action the reader should take in the first two sentences. 4. **Disclose when it matters.** If AI did the heavy lifting on something a client or leader will rely on, say so. 5. **Rewrite, don't forward.** A draft is a starting point, not a deliverable. Then, when a weak draft arrives, do the one thing that breaks the cycle: send it back with a specific note instead of fixing it silently. That one conversation teaches more than a month of quiet corrections. ## The Bigger Idea For two years the question was "how fast can we adopt AI?" The next question is quieter and harder: **how do we make sure what AI helps us produce is actually worth someone's attention?** The organisations that answer it will not be the ones with the most tools. They will be the ones whose people know what good work looks like, and are willing to hold each other to it. ## Frequently asked questions ### What is workslop? Workslop is AI-generated work content that looks polished but lacks the substance to meaningfully advance a task. The term was introduced in 2025 by researchers at BetterUp Labs and the Stanford Social Media Lab. Typical examples include a report with unverified figures, an email that sounds professional but makes no clear request, or a summary that misses the point of the source material. ### How common is workslop at work? In the BetterUp and Stanford survey of U.S. desk workers, 40% said they had received workslop in the previous month, and respondents estimated that about 15% of the work they receive fits the description. These figures come from self-reported surveys, so they show perception rather than audited counts, but they match what many managers describe. ### How much does workslop cost a company? The researchers estimated about two hours of rework per incident and about $186 per employee per month, which scales to roughly $9 million a year for a company of 10,000 people. These are estimates based on self-reported time. The less visible cost is lost trust: many recipients said the sender seemed less capable and less reliable afterwards. ### How can companies reduce workslop? Combine three things: training employees to use AI with judgment (AI literacy, verification habits), strengthening the writing and critical thinking skills that act as a quality filter, and giving managers a clear, shared standard for AI-assisted work. A short team agreement on ownership, fact-checking and disclosure is a useful first step, and structured training makes it stick. ### Is AI training widely available to Indian employees? Not yet at the depth most teams need. A Genius HRTech and Digipoll survey of 1,704 professionals, reported in January 2026, found that 61% said their organisation had not provided adequate guidance on using AI and only 37% had received proper training. A separate Salesforce India survey found that 94% of desk workers wanted to build AI skills, yet 40% had spent under five hours learning to use the technology. ### Which Bodhih programs help with AI quality at work? [AI Literacy Training](https://bodhih.com/ai-literacy-training/) builds safe, verified AI use across a workforce. [Business Writing with AI](https://bodhih.com/business-writing-with-ai-training/) and [Critical Thinking Training](https://bodhih.com/critical-thinking-training/) strengthen the judgment that filters AI output, and [AI for Leaders](https://bodhih.com/ai-for-leaders-training/) helps managers set quality standards for their teams. All can be customised for corporate batches.