Artificial intelligence is no longer just a tool that assists workers. It is increasingly becoming a system that coordinates, supervises, and in some cases directs work itself. Across technology companies and corporate offices, a new class of software known as AI agents is beginning to reshape how organisations operate and how employees interact with technology.
At Snowflake, the cloud data platform, engineers who once spent hours monitoring dashboards or gathering internal data now rely on AI agents to perform much of that groundwork. According to Qaiser Habib, the company’s head of Canada engineering, engineers interact with multiple AI agents each day to review product designs, assist during system outages, and even help write code. The result is a shift in the role of engineers themselves — away from routine operational work and toward higher-level decision making and oversight of automated systems.
Unlike traditional chatbots that simply respond to prompts, AI agents are designed to plan, reason, and execute multi-step workflows. These systems can draw information from calendars, internal databases, and meeting transcripts to complete tasks with limited human involvement. In some organisations, agents are even beginning to manage each other. One system may generate software code while another reviews it for errors and suggests improvements before a human approves the final version.
The implications for workplace structures are significant. Companies are already experimenting with AI systems that monitor performance, recommend promotions, assign tasks, and identify roles that could be automated. As organisations pursue greater efficiency, layoffs across several large corporations have been partly linked to automation initiatives. Goldman Sachs estimates that 6 to 7 percent of U.S. workers could eventually be displaced by AI, particularly in roles such as programming, accounting, administrative work, and customer service.
Yet the impact may be felt most strongly in the middle of corporate hierarchies. Experts increasingly believe that AI will not simply replace entry-level workers but will instead compress layers of management. Systems that translate executive strategy into operational tasks can now distribute assignments directly to employees. In this emerging model, workers increasingly supervise the systems that once performed their own jobs — becoming what some analysts describe as “managers of their previous roles.”
Despite the rapid pace of experimentation, the transformation is still in its early stages. Surveys show that many companies have yet to see measurable productivity gains from AI investments. Trust also remains a major barrier. Employees are often willing to delegate tasks to machines, but they may resist systems that threaten their autonomy or professional identity. Organisations that frame AI purely as a cost-cutting tool risk undermining cooperation and slowing adoption.
“We are open to revisiting AI systems in the future,” said New Ground Wellness cofounder Lucinda Bibbs, “but at this stage, preserving human connections remains our highest priority.”
What is becoming clear, however, is that the next phase of the AI revolution will not be defined only by smarter tools but by new organisational models built around human-AI collaboration. In that world, the most valuable employees may not be those who perform tasks the fastest but those who can effectively guide and manage the intelligent systems performing them.
Source: Al Jazeera – “The new boss at work may not be human” (March 8, 2026).
https://www.aljazeera.com/economy/2026/3/8/the-new-boss-at-work-may-not-be
