Companies are increasingly naming AI agents like Alex to mimic human coworkers. This anthropomorphism masks the tool's technical limitations and creates false expectations of reliability. Practitioners should treat these systems as software, not staff. Mislabeling agents as colleagues obscures accountability when the underlying LLM inevitably fails during complex tasks.