Technology

Jordan Belfort Launches Tone, a Voice AI Platform for Sales Calls and Audiobooks

Martin HollowayPublished 14m ago3 min readBased on 1 source
Reading level
Jordan Belfort Launches Tone, a Voice AI Platform for Sales Calls and Audiobooks
Image by AndNowProjekt from Pixabay

Jordan Belfort is building Tone, a voice-based AI platform, with longtime investor Shervin Pishevar as co-founder. The company's announcement, reported by TechCrunch on November 16, 2026, covers two product directions: sales outreach and audiobook-style narration.

On the sales side, Belfort says companies already using Tone have more inbound and outbound leads than their human teams can work. He also says Tone's AI voices draw on sales tonality techniques, including phrasing declarative statements as questions and whispering at times. Tonality here means the pitch, rhythm and emphasis of speech. For anyone who has built a dialer workflow or tuned a call script, the design claim is specific: prosody, the musical shape of a voice, is being treated as a conversion lever, not only as a target for sounding natural.

On the creator side, Tone is launching a self-service platform that lets content creators and authors turn books into expressive AI-generated narration. Users can pick from Tone's stock AI voices. A voice-cloning feature, which would let a user create a clone of their own voice, is rolling out. The announcement describes these as plans in progress, so cloning should be read as not yet generally available across the product.

Belfort's own record is part of the coverage. In 1999 he pleaded guilty to securities fraud and money laundering tied to Stratton Oakmont. He served 22 months in prison and was ordered to pay $110.4 million in restitution to 1,513 victims.

The broader context here is that the same person now fronts a company selling synthetic voices tuned for persuasion. Prospective customers, and the listeners on the receiving end of AI-placed calls, will weigh that history differently, and the company's pitch does not ask them to set it aside. The more concrete question for a buyer is how a sales-tuned voice is disclosed on a call, and whether the recipient can tell they are speaking to a synthetic agent. The announcement does not answer that.

Voice cloning raises a separate and familiar set of issues. A user who clones their own voice can extend their presence across many narrations or calls. The same capability pointed at someone else's voice is what consent frameworks and platform policies are built to prevent. Any cloning feature will need controls over who may create a clone and how a clone is verified, and those controls will shape how the feature is used in practice.

In my view, the authors' platform is the more consequential part of the announcement for the industry. Professional narration has historically been expensive and slow to schedule. A self-service path from manuscript to expressive narration could put production tools within reach of independent writers who could not afford them before. If the quality holds across genres and the cloning controls are sound, a working author could publish an audio edition on the same day as the text. That is the optimistic reading, and it depends on execution the announcement does not yet demonstrate.

Taken together, the facts are a product direction with two distinct customer groups, a sales-oriented voice design that its founder describes openly, and a creator tool closer to a general-purpose narration service. Customers and regulators will test each of those threads in ways a launch announcement cannot settle.