Former Meta AI Researcher Launches Legal Tech Startup
Ethem Coban is developing tools to help law firms build and customize their own proprietary artificial intelligence models.
Updated on Oct. 22, 2026 in Artificial Intelligence
Live Poll
Do you trust specialized in-house AI models more than software licensed from large technology providers?
After nearly two years at Meta's Superintelligence Labs, Ethem Coban has left the company to launch a startup focused on legal technology. His new venture aims to help firms customize AI models using their own internal case files and expertise.
Why it matters
Law firms are increasingly seeking to maintain control over internal data by developing their own AI models rather than relying on third-party systems. This shift allows firms to treat their vast archives of case files as a proprietary competitive advantage.
Ethem Coban helped train models during his tenure at Meta to improve legal question accuracy and reduce hallucinations. Kirkland & Ellis has committed $500 million to building proprietary software, while Latham & Watkins is purchasing Nvidia servers for model customization.
The players
Ethem Coban
He is a former researcher at Meta who spent nearly two years working at the company's Superintelligence Labs before launching a legal tech startup.
Meta
It is a major technology company that operates Superintelligence Labs and focuses on advancing artificial intelligence research.
Kirkland & Ellis
It is a global law firm that has committed $500 million toward building proprietary software to enhance its legal operations.
Latham & Watkins
It is a law firm that is investing in its own infrastructure by purchasing Nvidia servers to support in-house AI model customization.
The details
The startup focuses on building tools that enable law firms to refine AI models specifically around their internal legal expertise. Coban, who studied law at American University and Goethe University, is currently in discussions to secure $6 million in startup funding.
Timeline
October 2026: Ethem Coban left his position at Meta after nearly two years.
The Tech Race
This venture highlights the accelerating race among elite law firms to internalize AI capabilities rather than outsourcing to generic platforms. It marks a broader shift where legal data is being treated as the core intellectual property and fuel for custom model training.
Legal clients may eventually benefit from faster case resolutions as firms deploy more accurate and reliable AI tools. These developments could also change how law firms bill for services by automating complex research tasks that previously required extensive human hours.
The takeaway
The legal industry is rapidly pivoting toward owning its AI infrastructure to ensure data privacy and maintain a competitive edge. Firms and tech professionals looking to enter this space should prioritize specialized model training over general application deployment.
Further reading
Learn more about the latest innovations in the Artificial Intelligence sector.
Source note: This article includes information reported by Business Insider.
Live Poll
Do you trust specialized in-house AI models more than software licensed from large technology providers?
