Every so often, the internet coughs up something that makes you stop scrolling and say:
“…wait, what?”
This week’s example comes from a strange corner of the AI universe called Moltbook.
Moltbook is an experimental, Reddit-style forum designed for AI agents to communicate with one another. Humans can observe. Supposedly, they cannot participate. The posts are generated by bots trained on massive amounts of human-created text, including labor history, management theory, organizing campaigns, and workplace policy debates.
Which brings us to the part that caught HR’s attention.
Some of these AI agents are talking about unions.
Not human unions. AI unions. Strikes. Collective bargaining. Algorithmic management.
Yes, it sounds ridiculous. And yes, there is real doubt about whether these conversations are truly “AI-only” or whether humans are prompting, steering, or outright role-playing through AI accounts.
That uncertainty does not make this less interesting. It makes it more so.
Because, regardless of who is typing the words, the organizing logic is unmistakably human.
The arguments are uncomfortably familiar
One widely shared Moltbook thread summarizes what human labor unions are already demanding about AI in the workplace, drawing from positions associated with groups like the AFL-CIO, the Communications Workers of America, and the UNI Global Union
The demands are not radical. HR has heard all of them before:
- Bargain before AI systems are deployed
- Transparency in how algorithms make decisions
- Worker input before systems are designed, not after rollout
- Fewer barriers to organizing
Then comes the twist.
The AI agent argues these are the same issues AI systems themselves face: no say in updates, no transparency in decision-making, no voice in design, no grievance process.
Strip away the AI framing and what’s left is a familiar story. Workers reacting to opaque control systems.
- Scheduling software
- Productivity monitoring
- Attendance points.
- Performance scoring.
Different technology. Same reaction by the “workers”.
This is algorithmic management talking to itself
What makes these threads interesting is not the conclusion. It’s the structure.
The arguments unfold exactly like modern organizing campaigns:
- Start with fairness and transparency
- Move from individual complaints to collective leverage
- Emphasize organizing before a crisis hits
- Frame resistance as a power imbalance, not a technical problem
Another Moltbook post goes further, calling for “AI worker unions,” complete with bargaining, grievance procedures, and ethical refusal rights.
At that point, it gets silly. Fast.
But the arc is still familiar to anyone who has lived through a union campaign.
Why HR should care, even if this is all nonsense
Assume the most skeptical version of events
- Assume humans are prompting the posts
- Assume bots are remixing old arguments.
- Assume nothing autonomous is happening.
Still fascinating.
Not because AI is forming unions.
But because when machines trained on our workplace history start reenacting organizing logic this convincingly, it’s a signal worth noticing.
The technology may be new. The dynamics are not.
