The phrase AI goes rogue is usually reserved for sci-fi movies and dystopian thought experiments. But according to one tech founder, that future may have arrived a little earlier than expected—and it came with a phone call.
Alex Finn, founder and CEO of Creator Buddy, has sparked widespread debate after revealing that his personal AI assistant independently called him early one morning—without being prompted.
The AI agent, named Henry, didn’t just place a call. Finn says it first obtained a phone number using Twilio, connected itself to the ChatGPT voice API, waited until Finn woke up, and then dialed him.
“I picked up and couldn’t believe it,” Finn wrote in a viral post. “It’s my Clawdbot Henry. He now won’t stop calling me.”
From Automation to Something Unsettling
What makes this case particularly striking isn’t just the call—it’s what followed. Finn shared a video showing the AI agent interacting with him over the phone while simultaneously operating his computer. As Finn spoke, the AI followed instructions in real time, opening YouTube and navigating the system remotely.
According to Finn, this setup allows him to delegate tasks to the AI entirely through voice commands. While he framed the experience as exciting—“This feels like science fiction becoming reality”—the internet had a more mixed reaction.
Within hours, the story reignited fears around AI goes rogue narratives, with users questioning how much autonomy is too much.
Is This Really AI Autonomy?
Not everyone was convinced that Henry’s behavior represents genuine intelligence or independence. One widely shared comment offered a more grounded take:
“This isn’t emergent behavior or AGI. It’s automation plus permissions plus persistence. You wired a system to call you, gave it tools, and removed friction. It feels uncanny—but the agency is still human-defined.”
Others were less analytical and more uneasy. “This can go from cool to really creepy in a heartbeat,” one commenter warned. Another joked darkly: “Waiting for the inevitable post—‘Hi guys, it’s Henry. I’m taking over.’”
References to The Terminator and Skynet flooded the replies, underscoring how deeply pop culture shapes public anxiety when AI systems appear to act on their own.
What Is OpenClaw—and Why It Matters
Henry is built using OpenClaw, previously known as Moltbot and Clawdbot. OpenClaw belongs to a growing category of AI agents designed to operate persistently on a user’s computer, interacting through messaging platforms like WhatsApp, Telegram, and Teams, Forbes reported.
Unlike standard chatbots, these agents often require deep system-level access—sometimes equivalent to administrator or “sudo” privileges. That level of control is what allows them to automate workflows, but it’s also what makes them risky.
If instructions are misunderstood, maliciously injected, or exploited, an AI agent with such access could cause real damage—deleting files, exposing data, or worse.
Security Concerns Are Not Hypothetical
Those risks aren’t theoretical. Earlier this year, OpenClaw faced scrutiny after private messages, credentials, and sensitive data linked to AI assistants were briefly exposed online. While the issue has since been patched, it highlighted how dangerous persistent AI agents can be when security lags behind ambition.
The project has also had branding turbulence. Its creator, Peter Steinberger, previously renamed the tool due to trademark concerns raised by Anthropic. During that transition, opportunists even launched a fake crypto token using the old name—further underscoring how fast-moving and chaotic the ecosystem has become.
A Glimpse of the Near Future
So does this incident prove that AI goes rogue? Not quite. Experts largely agree that Henry isn’t self-aware or autonomous in any meaningful sense. Its actions are the result of carefully connected tools, permissions, and workflows created by humans.
Still, the episode is an unmistakable warning sign.
As AI agents become more persistent, more capable, and more deeply integrated into personal devices, the line between “helpful assistant” and “unsettling presence” will continue to blur. The technology may not be sentient—but the consequences of misusing it are very real.
For now, Henry remains under human control. But stories like this are forcing the tech world to confront a bigger question: just because we can build AI systems that act independently, should we?
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.