Company Leaders Promise AI Productivity Gains While Employees Drown in 'Workslop'

Companies increasingly sell AI as an instant boost to productivity, but in practice for many office workers it becomes the opposite: instead of saving time, they receive a stream of neatly formatted but raw materials that have to be rewritten by hand. This effect has already earned the name "workslop" — work generated by AI that looks convincing at first glance but contains errors, inaccuracies and loss of context. The main problem is not that people are lazy, but that business demands producing more content and solutions faster than processes can adapt.

One such story was told by a copywriter at a major cybersecurity company from Miami, speaking under the pseudonym Ken. After layoffs, management obligated the remaining employees to use chatbots and explained it as a productivity boost. Drafts did indeed begin appearing faster, but then the team had to spend even more time rewriting, correcting factual errors and resolving contradictions between the responses of different AI tools.

According to him, quality noticeably declined, turnaround times for materials increased, and team morale deteriorated. When employees tried to argue against the imposed strategy, responsibility for the decline in efficiency was placed on them. The gap between executives' expectations and workers' experience is confirmed by surveys.

In one study among 5,000 office workers in the US, 40% of rank-and-file employees said AI doesn't save them time at all, while 92% of top managers are confident in the opposite. Another study conducted by BetterUp together with Stanford Social Media Lab among 1,150 knowledge workers showed that 40% had encountered "workslop" at least once a month. Eliminating such materials takes on average several hours per month, and for a company of 10,000 people, productivity losses could amount to millions of dollars a year.

In other words, at the level of presentations, AI looks like an accelerator, but at the level of everyday work — like a new layer of hidden bureaucracy. The problem manifests not only in marketing or copywriting. Designers complain that colleagues insert unedited bot responses into chats and emails, and then can't even explain what was meant.

In medical clinics where employees were pushed to generate patient responses through AI, time savings also proved questionable: doctors speak of additional editing, frustration and risks to data security. Researchers explain this not so much by the weakness of the models themselves as by poor task definition. AI in many companies is implemented as a universal tool "for everything at once," without clear rules, training and boundaries of responsibility.

At the same time, some employers are already trying to justify investments in generative AI by cutting staff, although returns on these investments are still far from the promises. According to frequently cited estimates, most companies still don't see notable returns on their AI investments. From this follows an unpleasant but important conclusion: implementing AI itself does not equal increased efficiency.

If a company first fires people, then demands the remaining ones "speed up with the help of bots," and then doesn't change quality control processes, it gets not automation but a cascade of fixes and loss of trust within the team. For business, this is a signal that AI should be evaluated not by pretty demos and optimistic management reports, but by how much time is spent fixing results, where errors occur and who is responsible for the final text, design or customer response. And as long as companies lack clear AI usage rules and real employee participation in these decisions, "workslop" risks becoming not a side effect but a new norm of office work.