Tech companies are aggressively laying off workers to invest in artificial intelligence agents that research shows cannot reliably perform the jobs they're meant to replace. A major AI infrastructure company found that even the best AI agents complete tasks to professional standards less than five percent of the time, yet Meta cut nearly ten percent of its workforce last month, Block slashed its staff almost in half in February, and Microsoft and Amazon have laid off thousands citing shifts toward AI.

How Bad Are AI Agents at Real Work?

San Francisco-based Scale AI, which provides data and evaluations to governments and Fortune 500 companies, developed a benchmark called the Remote Labour Index to measure how well AI agents can actually perform real-world, economically valuable remote work from start to finish. The company tested agents on tasks sourced from professional freelance platforms like Upwork, including designing leaflets and logos, editing and generating videos, and developing architectural models, then compared the results to work produced by human freelancers.

The findings were sobering. According to Scale AI's research, even the best-performing AI agents succeeded at producing client-ready work less than five percent of the time. When the company started its benchmark in fall 2025, the top-scoring agent had a success rate of just 2.5 percent. By March 2026, that improved to 4.17 percent, showing progress but revealing a long road ahead.

The most common reason for failure across all projects was poor quality, described as "child-like" or amateur work, accounting for about forty-six percent of failures. More than one-third of failed submissions were incomplete, about eighteen percent were corrupt or used incorrect file formats, and about fifteen percent "failed to maintain visual or logical consistency across files".

Where Do AI Agents Actually Succeed?

AI agents showed their highest success rates in specific, narrower tasks. The technology performed best in image generation, report writing, audio tasks, and data retrieval. However, agents failed most commonly in complex tasks requiring detailed specifications, such as producing architectural drawings based on particular client requirements.

A separate study by researchers at Stanford and Carnegie Mellon universities, published in October 2025, found that while AI agents work faster and cheaper than humans, they "produce work of inferior quality" and mask their deficiencies with "data fabrication and misuse of advanced tools".

Steps to Understanding the Gap Between AI Hype and Reality

• Benchmark Performance: Scale AI's Remote Labour Index shows AI agents succeed at professional-quality work less than five percent of the time, compared to human freelancers who consistently meet client standards.
• Quality Issues: Nearly half of AI agent failures stem from poor quality described as amateur or child-like, while one-third produce incomplete work and fifteen percent fail to maintain consistency across files.
• Task Complexity: AI agents perform better on straightforward tasks like image generation and report writing but struggle with complex, multi-step work requiring specific specifications and creative problem-solving.
• Improvement Rate: AI agent success rates improved from 2.5 percent to 4.17 percent over six months, indicating slow progress on complex, holistic tasks that require human-level reasoning.

Are Companies Using AI as a Layoff Excuse?

Despite the poor performance data, tech companies continue announcing AI-related workforce reductions. Outplacement services company Challenger, Gray and Christmas found that AI led all reasons for U.S. job cuts in March and April 2026. In May, research firm Gartner surveyed 350 global business executives with annual revenue of at least one billion dollars and found that eighty percent of those who piloted an AI or autonomous technology reduced their workforce due to automation.

However, some industry leaders acknowledge the technology is not ready for the hype surrounding it. Scale AI CEO Jason Droege told Semafor World Economy in April that corporate customers often ask for automation to generate savings and productivity, but he steers them away because "there's a lot of problems that the technology's not mature enough to solve with reliability and safety." Droege added that some companies are "washing" their layoffs, using AI as an excuse to reduce their workforce.

Jason Droege

"AI requires a lot of money. It's like a cash-sucking experiment, and no companies can really predict if it will work, so they have to lay off workers to keep costs low. Sometimes using technological unemployment is just an excuse," said Julie Yujie Chen, an associate professor at the University of Toronto who studies how digital technologies are transforming work.

Julie Yujie Chen, Associate Professor, University of Toronto

OpenAI CEO Sam Altman similarly acknowledged in February that while some AI displacement is real, some companies are "AI-washing, where people are blaming AI for layoffs that they would otherwise do".

Interestingly, companies often see their stock prices jump after announcing AI-related layoffs. Block's stock spiked more than twenty percent following its announcement, though such gains can be short-lived. Chen argues that the investment in AI as storytelling has more value than keeping a workforce in the current economic system.

What Happens to Workers Who Remain?

Several experts told CBC News that the agentic AI trend will lead to an "intensification" of work for humans who remain employed. Their jobs will likely shift to reviewing, verifying, and monitoring multiple AI agents rather than doing the work themselves. Meta appeared to outline this vision in a memo obtained by Reuters in April, in which the company's Chief Technology Officer Andrew Bosworth told employees that Meta is building toward a future "where our agents primarily do the work and our role is to direct, review and help them improve." The memo also revealed that Meta would begin surveilling employee keystrokes and mouse movements to train AI agents.

A Meta spokesperson rejected the claim that agentic AI is being used as an excuse for layoffs, stating in an email to CBC News that "Conflating this as a tool for layoffs would be wildly inaccurate".

The disconnect between AI agent capabilities and corporate investment decisions reflects a broader tension in the technology industry. While research demonstrates that autonomous software programs using large language models still lag far behind human workers on complex, multi-step tasks, the narrative of AI-driven transformation continues to drive major business decisions and workforce reductions across the sector.