Artificial intelligence agents—autonomous systems based on large language models—are rapidly being deployed as digital workers in corporate environments, yet organisations are largely unprepared for this shift. An estimated 22 per cent of surveyed organisations have already added AI agents to their corporate structures, with this figure expected to grow as technology companies and startups aggressively market these services as the ultimate autonomous employees. The lack of clear terminology and frameworks for how these agents should function within teams has created confusion about whether they should be viewed as tools, teammates, or full colleagues.

Companies are deliberately humanising their AI agents through strategies such as assigning them names, roles, cloud-based "identities," and Muppet-esque avatars to encourage employees to accept them as coworkers. Agents like Microsoft's Scout and platforms such as Skydive are being deployed for tasks ranging from scheduling meetings and drafting emails to writing contracts and monitoring systems around the clock. Remarkably, employees are already forming attachments to these digital colleagues, referring to "working with" specific agents and trusting them with sensitive workplace communication—whilst also finding it easier to be blunt with them than with human colleagues precisely because they lack human feelings.