The Download: AI 'Coworkers' and Stratospheric Internet
Plus: The US House has passed new youth online safety legislation.
Welcome to today's edition of The Download, our weekday newsletter bringing you the latest developments in technology. Here’s what you need to know for June 30, 2026.
AI Agents Are Not Your 'Coworkers'
Imagine arriving at work to discover a new team member reporting to you—except this “employee” is not a person but an AI tool. Your company calls it Alex, assigns it a title and defined responsibilities, and expects you to collaborate. How effectively would you work with Alex?
If your experience mirrors that of managers studied by Boston University professor Emma Wiles, the answer is: not well. Treating an AI as a “coworker” rather than a simple chatbot can impair performance. In her research, managers who believed an error-prone task came from an agentic “AI employee” caught 18% fewer mistakes than those who thought the same work was from a chatbot. The label alone—framing AI as a colleague—led to reduced vigilance.
This finding offers a troubling preview of the future Silicon Valley is rapidly building. By 2026, Microsoft, OpenAI, Anthropic, and Google have all released tools for managing teams of AI agents, many marketed as digital colleagues. Yet as James O'Donnell explains in his latest feature, treating AI as a peer is a losing proposition for workers. When companies frame machines as coworkers, humans may lower their guard, trust automation too much, and miss critical errors. For a deeper dive into why this approach backfires, read the full article: AI Agents Are Not Your Coworkers.
—James O'Donnell