In detail
- Managers treating AI agents as 'coworkers' missed 18% more errors than when framed as chatbots.
- 44% more likely to escalate questionable work to a manager rather than correct it themselves—negates time savings.
- Nearly one-third of 1,261 surveyed managers say their companies frame AI agents as employees; 23% list them on org charts.
- Microsoft, OpenAI, Anthropic, and Google have since April released tools for 'digital colleagues' with flexibility of actual humans.
Why it matters
This is a critical warning against anthropomorphization: when AI agents are framed as employees, human accountability and quality control decline. For German companies, this is central—false expectations and accountability vacuums can lead to poor decisions.
For you Clearly define AI agents as tools, not colleagues—and establish explicit control processes to prevent accountability gaps.