AI tools shift more work from employees to users, report argues

In short: A growing view is that AI is moving work not just from workers to machines, but also from workers to you.

What's going on

Some writers and researchers say the “AI job shift” is being misunderstood. The common story is that software does the work that people used to do. A newer argument says there is also a transfer of effort from workers to end users, meaning regular people get useful results while more of the human work is hidden or pushed onto them.

One part of this is “ghost work.” These are often low paid workers who label training data (they tag things so the system can learn, like putting name stickers on thousands of photos) and moderate content. A 2025 to 2026 report from the Communications Workers of America describes how large tech companies often rely on layers of subcontractors, which can make the workers less visible to the public. The report gives examples where services that look automated are still supported by remote human help.

Another part is that AI can remove tasks piece by piece. When that happens, companies may hire fewer entry level workers, but the remaining work does not vanish. Users may end up doing more setup, checking, and correction, like writing careful prompts (instructions) and verifying the output before it is safe to use.

What to watch

Watch for how companies explain what is “automated,” and who is still doing the behind the scenes work. Also watch whether more products require users to spend time learning prompts and checking results, and whether workers who support AI systems get clearer pay, protections, and credit.

Source: NYTimes