You see a cute, blinking robot rolling into a New York classroom. It’s designed to help kids learn, smile, and feel comfortable. It looks like a harmless piece of tomorrow’s tech. But if you look at the company that built it, you might want to unplug it immediately.
This was the reality for a school in Salamanca, New York, when they invited an educational robot into their halls—only to pull the plug when they discovered who its siblings were. The robot was built by Realbotix, a company that also manufactures AI-powered sex dolls with names like Serenity and Harmony. Yes, the very same underlying technology, sensors, and software that’s supposed to teach your kid math is just a different shell away from being the “world’s most advanced sex robot.”
AI doesn’t have a moral compass; it has a product portfolio.
We like to compartmentalize technology. We want to believe that the AI in our schools is pure, sanitized, and entirely separate from the AI used in adult entertainment or surveillance. But the invisible supply chain of artificial intelligence doesn’t care about your branding boundaries. The foundational models, the computer vision, the natural language processing—it’s all the same. The market’s reaction to the Sally the Robot debacle reveals our deep, collective unease with that flexibility.
When a company builds a benign educational tool and an intimate adult product under the same roof, it exposes the hidden ethical risks every AI company faces. It forces parents, educators, and investors to look beyond the plastic shell. You aren’t just buying a teaching assistant; you are inviting a company’s entire ethos into a room full of children.
The same sensors that teach your kid the alphabet are just a software patch away from a very different kind of ‘intimacy’.
Most discussions about AI in schools focus on performance: Can it teach? Will it replace teachers? But the real story is about corporate identity. If the company making your child’s robot is primarily motivated by the adult market, their data priorities, safety protocols, and future updates are inherently compromised. You cannot separate the creator from the creation when the creation is just a repackaged version of the creator’s other products.
This isn’t just a PR blunder for Realbotix. It’s a warning shot for the entire AI industry. As artificial intelligence scales, dual-use applications are the rule, not the exception. We have to stop evaluating AI tools in a vacuum. We must demand transparency not just about what the robot does in the classroom, but about what its parent company does in the dark.
If you don’t know what else your AI vendor is building, you don’t know what you’re letting into your home.
FAQ
Q: Isn't it unfair to judge an educational robot by what its sibling company makes?
A: No. When the underlying technology, sensors, and software are shared, the corporate priorities bleed over. You can't separate the data practices of an adult toy manufacturer from the data practices of a classroom robot if they share a brain.
Q: What should schools do before adopting AI tools?
A: Look past the product. Audit the parent company's entire portfolio, data policies, and revenue streams. If a vendor's main business is in adult entertainment or aggressive data harvesting, their educational product is just a Trojan horse.
Q: Is all dual-use AI inherently bad?
A: No, but it is inherently dangerous. The fact that the same tech can teach a kid or provide 'intimacy' proves AI is just a tool. The danger lies in pretending the tool's context doesn't matter. Own the dual-use reality, or get blindsided by it.