AI · 6 October 2026
Ghost Raises $11M for $3,499 Core AI Agent Computer
Startup Ghost, founded by a 19-year-old entrepreneur, has raised $11 million to build Core, a dedicated $3,499 device designed to run autonomous AI agents.
What happened
Ghost, a startup founded by a 19-year-old entrepreneur, has raised $11 million to build Core, a personal computer designed specifically to run AI agents capable of taking actions on a user's behalf. The device is priced at $3,499 and is being positioned as dedicated hardware for personal AI, distinct from general-purpose computers running AI software on top of existing operating systems.
According to reporting from TechCrunch, Core is Ghost's first product and represents a bet that AI agents — software capable of completing tasks autonomously rather than simply answering queries — will eventually need purpose-built machines rather than running as an add-on to conventional laptops or desktops.
Why it matters
The funding and product signal a broader shift in how some founders and investors think AI will be consumed: not as an app layered onto existing devices, but as the organising principle of the hardware itself. If agentic AI becomes a genuine computing paradigm rather than a feature, there's a case for machines engineered around delegation and autonomous action from the ground up — much as smartphones were built around mobile-first software rather than retrofitted desktop operating systems.
For technology and experience leaders, Ghost is an early signal worth watching rather than an established market. A dedicated "agent computer" raises real questions about trust, permissions and control — how much autonomy users are comfortable granting a device that can act for them, and what service and support models are needed when something goes wrong with an action taken on a person's behalf.
By the numbers
- $11 million raised by Ghost to develop Core
- $3,499 price point for the Core device
- 19 years old — the age of Ghost's founder at the time of the raise
The Renascence take
Hardware bets on agentic AI are as much a behavioural experiment as a technical one: the real product being sold is trust, delegated in advance, at a price point that assumes the buyer is ready to hand over meaningful autonomy to a machine.
A $3,499 price tag is itself a design choice — it filters for early adopters willing to tolerate friction in exchange for status and novelty, which buys Ghost time to learn before mainstream scrutiny arrives. The harder problem isn't building an agent that can act; it's designing the moments where a user needs to interrupt, correct or revoke that action without feeling like they've lost control of their own computer. Any operator building agentic experiences, on dedicated hardware or not, should treat the "undo" and "explain yourself" interactions as the real product, not an afterthought bolted on once the autonomous features are working.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in AI
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.
