The Trillion-Dollar Industries AI Is Disrupting: Voice, Law & the End of the Billable Hour

ElevenLabs was founded in 2022 and reached 100 million dollars in ARR in roughly 20 months, then accelerated to 200 million in 10 more months and 300 million in 5 more months. The company is now at 600 million dollars in revenue with 600 employees and zero attrition from its original first 10 employees, making it one of the fastest enterprise companies with a direct sales motion to grow from one to 150 million dollars in a single quarter.

The company describes itself as a communication platform for AI covering speech generation, transcription, and orchestration. It has never hired product managers and instead embeds engineers across every team including talent, legal, and go-to-market functions, where those engineers both build automations and serve as a security check on AI-generated code deployed by non-engineers who may not be able to evaluate whether what they build is secure. Enterprise sales to financial services clients including Revolut, Klarna, and Pagbank represent the biggest driver of recent growth. Users have been found to disclose sensitive financial information more willingly to an AI voice agent than to a human due to reduced shame, and legacy IVR systems were so disliked that AI voice agents have now improved to the point where some users feel guilty routing simple tasks to a human agent.

ElevenLabs has paid over 22 million dollars back to its community of voice talent through a marketplace where actors can record for an hour, create a licensed voice, and earn ongoing revenue. The company powered the Darth Vader voice in Fortnite by Epic Games following a deal James Earl Jones made with Disney to license the voice for all time, and it partnered with Matthew McConaughey on a World Cup voice experience across multiple languages. MasterClass uses ElevenLabs to create interactive AI instances of talent such as Gordon Ramsay. The company has also worked with ALS and throat cancer patients to restore lost voices, and Congresswoman Jennifer Wexton used ElevenLabs to deliver the first AI-assisted speech in Congress. On safety, the platform prevents voice cloning without the subject's involvement, traces all generated content, moderates at both voice and text level to block commercial scam attempts, and built a detection system allowing anyone to upload a sample and learn immediately whether it is AI-generated.

ElevenLabs operates as a model-agnostic platform offering customers access to OpenAI, open source, and Google models, which it describes as a competitive advantage because customers can build agent orchestration without dependency on any single provider. The company pays tens of millions of dollars annually to frontier model providers and is exploring building its own language model capability focused on knowledge work and coding. Data leakage from frontier model partners is described as an existing problem with only mechanisms to slow it down rather than stop it. On the research side, the company argues architecture matters more than scale for voice model performance.

Legal services represent a trillion-dollar annual market that is highly fragmented, yet legal technology software spend is only approximately 40 billion dollars, or about 4 percent of total legal services spend. Kirkland and Ellis generates approximately 10 billion dollars in annual revenue with 4,000 to 5,000 lawyers and partners earning between 5 and 10 million dollars per year. Senior partner billing rates range from 1,800 dollars per hour up to 4,000 dollars at firms like Kirkland. Founders and lawyers operate with structurally misaligned incentives, with founders motivated to close deals quickly while lawyers are motivated to avoid liability and maximize fees. One company conducted acquisition due diligence using its own tool and closed a transaction in 12 days from LOI to closing, and a seed-stage startup at one million in revenue used ChatGPT instead of a corporate lawyer for contracts, cap table, HR, and IP assignments.

AI disruption of legal services is categorically different from prior technology shifts like document management software because AI can perform a large portion of the substantive legal work itself rather than offering a mild productivity gain. Anthropic Opus 4.5 and 4.6 releases enabled agents to perform intelligent case strategy by combining witness statements and case materials end to end, marking a shift from AI augmenting human legal work to AI actually performing it. Legora and Harvey combined are already generating hundreds of millions of dollars in revenue, directly threatening LexisNexis, whose stock is being negatively repriced due to AI disruption expectations. The early market assumption that legacy data holders would win is proving incorrect, as those providers struggle to attract talent and move at the pace required. Building a credible legal research solution requires access to all case data rather than just the top 80 percent, because litigators at top firms cannot risk missing a single case, and obtaining comprehensive US case law still requires physically visiting courthouses and paying per-page scanning fees. Getting AI products into law firm customers is described as very hard despite the technology being relatively easy to build, making existing trust inside a firm a meaningful competitive advantage.