In a recent LinkedIn post, Ben Eubanks raises critical questions about the unchecked adoption of Artificial Intelligence (AI) in professional settings, particularly concerning sensitive workforce decisions. Eubanks uses a striking anecdote about a man developing bromide poisoning after following ChatGPT’s advice to illustrate the potential dangers of blindly trusting AI-generated guidance.
He highlights the severity of such missteps, stating:
“A man developed bromide poisoning (AKA bromism) after following AI advice. Because he wasn’t a medical professional and didn’t know better. Apparently AI didn’t know better either.”
This cautionary tale serves as a springboard for Eubanks’s broader argument that while AI can be a valuable tool, it requires significant human oversight, especially in areas impacting people’s livelihoods and well-being.
The Perils of Unfettered AI in Workforce Management
Eubanks emphasizes that the consequences of AI errors can range from minor inconveniences to severe repercussions. He points to instances where AI has provided outdated or incorrect information, leading to wasted time and opportunities, such as a planned weekend getaway where a festival had been canceled – a detail the AI failed to verify.
However, the core of his concern lies in the application of AI to critical human resources functions. Eubanks questions the wisdom of using AI for decisions like hiring, determining compensation, or building succession plans without robust human intervention.
The Necessity of Human Judgment
According to Eubanks, the concept of “humans in the loop” extends far beyond simple prompt engineering. He argues that true oversight involves human judgment, discernment, grounding, understanding, forethought, and insight – qualities that AI, in its current form, cannot replicate.
“AI needs more governance than that. Humans in the loop is more than a person prompting a machine. It’s judgment and discernment. It’s grounding and understanding. It’s forethought and insight.”
Eubanks suggests that the risks associated with AI making independent decisions in hiring, retention, and development are too high. He poses a direct question to readers, challenging them to consider their comfort level if AI were making critical decisions about their own families.
Employee Comfort Levels with AI in Hiring
Further underscoring his point, Eubanks references survey data indicating a significant difference in comfort levels when AI is used to make decisions about others versus oneself. He notes:
“800 US workers were relatively comfortable with AI being used to hire OTHERS. But in a previous survey when we asked if they’d be comfortable with AI being used to hire them? Much less comfortable.”
This disparity highlights a common human tendency to be more accepting of AI’s role when the direct personal impact is perceived as minimal, a sentiment Eubanks urges readers to overcome.
Maintaining Control in an AI-Driven World
While acknowledging that AI is an inevitable and valuable addition to many professional tools, Eubanks concludes with a strong call for vigilance. He advises professionals to remain actively engaged and in control of critical decision-making processes.
“AI is here to stay. It has value in some areas. But the human tendency to let AI do its own thing is strong. Keep your hands on the wheel, friends.”
Eubanks’s post, inspired by a recent discussion with the Harver team, serves as a timely reminder for businesses to approach AI integration with caution, ensuring that human oversight remains paramount in areas where judgment and ethical considerations are essential.
📝 About This Content
This article is based on insights shared by Ben Eubanks on LinkedIn.
📅 Originally posted on July 24, 2026 | View original post on LinkedIn →