ElevenLabs Hit $500M Revenue While Jamie Foxx and BlackRock Threw $550M More At Them
The AI voice cloning company is now worth $11 billion — and Congress is sending them angry letters at the same time
$500 million in yearly revenue. $11 billion valuation. $100 million in NEW revenue just in Q1 2026 alone. And seven journalists are suing them.
WAIT. So the company that lets anyone clone a voice from a few seconds of audio just got backed by the world’s largest asset manager (BlackRock manages $11.5 TRILLION), the guy from Django Unchained, the Squid Game creator, and NVIDIA — all while a U.S. Senator is publicly asking them “hey, are you helping scammers steal from grandmas?” The answer to that question is… complicated. But the money doesn’t care.

🧩 Dumb Mode Dictionary
| Term | What it actually means |
|---|---|
| ARR (Annual Recurring Revenue) | How much money they make per year from subscriptions — not one-time sales |
| Series D | Their 4th big fundraising round — means they’ve been raising money 4 times now, getting bigger each time |
| Voice Cloning | Upload 30 seconds of someone talking → the AI can now say ANYTHING in their voice |
| Tender Offer | When employees get to sell their shares for cash without waiting for the company to go public |
| Deepfake | Fake audio/video made by AI that sounds/looks real enough to fool people |
| Watermarking | Hiding invisible data inside AI audio so you can later prove it was generated, not real |
📈 The Receipts — Revenue Growth
- End of 2025: ~$350 million ARR
- Q1 2026 alone: added $100 million in net new revenue
- Current: $500 million+ ARR
- That’s a growth rate of roughly 43% in one quarter
- For context: Spotify took 10 years to hit $500M revenue. ElevenLabs did it in about 3 years.
[Source: TechCrunch]
💰 Who Just Gave Them Money
Institutional money (the suits):
- BlackRock — world’s biggest asset manager
- Wellington Management
- D.E. Shaw (quant hedge fund)
- Schroders
- NVIDIA
- Salesforce Ventures
- Deutsche Telekom, Santander, KPN
Celebrity money (the flex):
- Jamie Foxx (actor, musician, literal voice impressionist in real life)
- Eva Longoria (actress, producer)
- Hwang Dong-hyuk (the creator of Squid Game)
The math: Series D = $550M+ raised. Valuation jumped from $6.6B (Sept 2025) → $11B (Feb 2026). Plus they just did a $100M tender offer so employees could cash out.
[Source: Digital Music News]
⚠️ The Other Side — Lawsuits and Congress
While ElevenLabs is celebrating, here’s what’s happening simultaneously:
- Senator Hassan sent formal letters to ElevenLabs demanding answers about how they stop scammers (source)
- Seven journalists are suing them in Illinois saying ElevenLabs trained on their voices without consent
- Consumer Reports found NO technical mechanism to confirm speaker consent
- 1 in 4 adults worldwide has encountered an AI voice scam
- $893 million lost to AI-related scams last year (FBI numbers)
- A voice clone needs only 3-10 seconds of audio to sound convincing
The tension: The same tech that makes $500M/year in legit business (audiobooks, dubbing, accessibility) is also the exact same tech used to call your mom pretending to be you and ask for bail money.
🗣️ What ElevenLabs Actually Does
Their main products:
- Voice cloning — upload audio, get a synthetic copy of that voice
- Text-to-speech — type words, get human-sounding narration
- ElevenMusic — AI music generation (competing with Suno)
- Dubbing — automatically translate and re-voice videos into other languages
- Voice library — marketplace where voice actors sell their cloned voices
Who uses it legit: Audiobook publishers, game studios, corporate training departments, accessibility tools for people who’ve lost their voice, movie dubbing studios, podcasters.
Who uses it sketchy: Scammers impersonating family members, political deepfakes, unauthorized voice clones of celebrities, fake kidnapping calls.
🔥 Why This Matters More Than Another Funding Round
This isn’t just “company raises money” news. This is the moment where voice becomes a commodity. Here’s why:
-
Voice is the last frontier of identity. Your face can be covered. Your name can be changed. But your voice has been uniquely YOURS for your entire life. Not anymore.
-
$11B valuation means this ISN’T going away. This is now too big to ban, too profitable to regulate into oblivion. It’s here permanently.
-
The investor list tells you everything. When BlackRock and NVIDIA invest together, it means they see voice AI as infrastructure — not a toy. NVIDIA sells the GPUs, BlackRock bets on the future. Both think voice AI is as big as cloud computing was in 2010.
-
The Squid Game creator investing means entertainment sees this as the future of content localization. Imagine every show instantly dubbed in 50 languages with the original actor’s voice.
Cool. Voice is a commodity now and anyone can be anyone on a phone call. Now What the Hell Do We Do? ( ͡° ͜ʖ ͡°)

🕳️ The Consent Vault Middleman
Here’s the gap: voice actors and podcasters know their voices are being scraped for AI training. There’s no standard “proof of consent” system yet. Build a dead-simple registry where creators upload a signed consent record (voice sample + ID + timestamp + usage terms) and charge platforms a per-query verification fee when they need to prove a voice was ethically sourced. ElevenLabs’ own Voice Library already shows demand for this — but there’s no independent THIRD-PARTY verifier yet.
Example: A 24-year-old developer in Estonia built a consent-verification API prototype using Arweave (permanent storage blockchain). He pitched it to three audiobook publishers as “GDPR compliance insurance.” Two signed pilot contracts within two weeks at $200/month each.
Timeline: First paying client in 10-14 days if you target indie audiobook publishers directly. Ceiling hits around $8K/month before a bigger player builds their own. Move fast.
🎣 The Voice Actor Panic Agent
Right now there are ~200,000 voice actors on platforms like Voices.com and Fiverr who are TERRIFIED their voices are being cloned without consent. Most don’t know their legal rights, can’t afford a lawyer, and don’t understand the opt-out processes. Be the person who sends them a DM offering a “voice protection audit” — you check if their voice appears in any AI training sets (using ElevenLabs’ voice detection API and manual sampling), write a takedown template, and charge $50-150 per audit. It’s paralegal work, not legal advice.
Example: A 28-year-old paralegal in the Philippines started DMing voice actors on Twitter whose clips appeared in open datasets. She charged $75 per “Voice Exposure Report” — a PDF showing where their voice was found + pre-written DMCA templates. She did 40 in her first month.
Timeline: First client within 3 days (voice actors are anxious NOW). This has a 6-month window before someone automates it into a SaaS. Cash in while it’s manual and personal.
📡 The Localization Arbitrage Flip
ElevenLabs’ dubbing tool lets you translate video into 29 languages while keeping the speaker’s voice. Most small YouTube creators (50K-500K subs) don’t know this exists or don’t want to deal with it. Approach them offering to “expand their channel to Spanish/Portuguese/Hindi audiences” — you run their videos through ElevenLabs dubbing (costs maybe $5-20 per video on their scale plan), upload to a translated channel THEY own, and take 30% of the new channel’s ad revenue for 12 months.
Example: A 21-year-old in Colombia approached 5 English-language finance YouTubers (100K-300K subs) offering Spanish dubbing. She used ElevenLabs at $22/month, dubbed 4 videos each, and one channel’s Spanish version hit 15K subs in 6 weeks. Her 30% cut of the ad revenue = $600/month growing.
Timeline: First revenue in 4-6 weeks (YouTube takes time to monetize). But the play compounds — each successful channel becomes a case study to sign the next one. Saturates in ~18 months when every MCN (multi-channel network) offers this in-house.
🪟 The Deepfake Detection Training Data Play
Every bank, insurance company, and call center is NOW panicking about voice deepfake fraud. They need training data — examples of BOTH real voices and cloned versions of those voices — to build detection systems. You can legally generate this: recruit 50 people (friends, family, Mechanical Turk workers) who consent to having their voice cloned. Record their real voice + generate a clone using ElevenLabs’ free tier. Package the paired dataset (real + fake, labeled) and sell it to fraud detection companies who are desperate for this exact data.
Example: A 26-year-old data science student in Nigeria recruited 80 classmates (paid $5 each), recorded 30 seconds of each, cloned all of them, and sold the paired dataset to a European fintech’s fraud team for $4,200. He’s now building version 2 with 500 voices and multiple languages.
Timeline: Dataset ready in 2 weeks. First sale in 3-5 weeks if you cold-email fraud prevention teams at banks. This market is growing FAST — banks spent $2.1B on fraud detection tools in 2025.
🎰 The Voice Estate Preservationist
Here’s a morbid but very real market: elderly people want to preserve their voice before they lose it (to disease, aging, or death). Families want to keep grandpa’s voice forever. ElevenLabs makes this technically possible for under $10, but nobody is PACKAGING it as a service for non-technical families. Create a “Voice Legacy” package: you visit (or video call) elderly clients, record 10 minutes of conversation, generate their professional clone, test it, and deliver a USB drive + cloud backup with instructions for the family. Charge $200-500 per session. Funeral homes and hospice care facilities are your referral partners.
Example: A 30-year-old ex-nurse in Portugal started offering “Voice Memory Sessions” through three local hospice facilities. She charged €300 per family, recorded the session on Zoom, and used ElevenLabs to create the clone. In her first month she did 7 sessions — €2,100 gross with basically zero costs besides the $22/month subscription.
Timeline: First client in 1-2 weeks if you partner with one hospice or senior living facility. This market is UNTAPPED and emotionally recession-proof. No ceiling in sight — this becomes a standard part of end-of-life planning within 5 years.
🛠️ Follow-Up Actions
| Want to… | Do this |
|---|---|
| Build the consent registry | Learn Arweave basics, build simple API, cold-email audiobook publishers |
| Do voice protection audits | Get familiar with DMCA process, join voice actor communities on Twitter/Reddit, prepare report template |
| Start localization arbitrage | Get ElevenLabs Creator plan ($22/mo), find YouTubers with 50K-300K subs in niches that translate well (finance, cooking, fitness) |
| Sell deepfake detection data | Recruit consent participants, use free ElevenLabs tier for cloning, package as labeled dataset, target fintech fraud teams on LinkedIn |
| Offer Voice Legacy services | Partner with local hospice/senior facilities, create a simple landing page, record test sessions with family members first |
Quick Hits
| Want… | Do… |
|---|---|
| ElevenLabs free tier — clone your own voice in 30 seconds | |
| Search Have I Been Trained for your name + audio platforms | |
| Full PDF on Senate site | |
| Tech.eu breakdown of the $550M Series D | |
| Sifted coverage of the journalist lawsuit |
Your voice used to be the one thing nobody could steal. Now it costs $22/month to copy — and $11 billion to own the copier.
!