Relativity CEO Confirms 2028 Cloud Deadline, Prioritizes Frontier Labs

Relativity is holding firm on its plan to require that all new matters be hosted in the cloud beginning Jan. 1, 2028, CEO Phil Saunders said Thursday, and he had a blunt message for partners who are not on board: Know were going to run through you. Saunders made the comments during a media briefing on the last day of Relativity Fest in Chicago, where he and Chris Brown, who was named president in July , took questions from reporters. Asked by me whether the company is sticking to the deadline it announced in January 2025 to require that all new matters be hosted on its cloud platform rather than its legacy Relativity Server, Saunders quickly answered, Yes. The policy is not to end the life of Server, he said. The policy is to make sure that no net new matter is going to Server effective Jan. 1. Saunders said he delivered the same message earlier in the week at a briefing for several hundred of the company’s partners – the companies and consultants who are central to distributing and implementing Relativity for legal, investigative and compliance teams. He described partners as falling into three buckets. The first are super duper aligned, already sending new workspaces and data to the cloud, adopting the companys AI capabilities, and collaborating with Relativity on the end customer. At the other extreme, he said, there are some that are “just not aligned. He attributed that to business models built around buying Server as a capital expense and depreciating it over time. In between are those we would politely call hybrids, doing some of both. We dont blame them for that, he said, noting that many of their customers remain on Server. Saunders said partners that ignore the deadline are serving their own pockets, not the end customer. He said that even Relativity earns less from the cloud. We make a lot less money on Relativity aiR than we do on Server, like a lot less, he said. But it doesnt matter. Its about disrupting yourself to service where the market and business is going. For partners on the fence, he said, Relativity has budgeted credits and money and effort to help them move from a capital-expense to an operating-expense model. The aligned partners are going to win, he said, and the non-aligned partners – let’s talk about it. Saunders said that as of the end of June, nearly 84% of the companys annual recurring revenue was in Relativity aiR, the name the company recently gave its cloud platform , formerly RelativityOne. When it announced the deadline in January 2025, the company put the figure at more than 75%. Saunders also said Relativity is prioritizing partnerships with the frontier AI labs over integrations with legal AI companies such as Harvey and Legora. A reporter at the briefing, Stephanie Wilkins, director of content at Legaltech Hub, brought up the fact that at last year’s briefing, Saunders said he was in conversations with the founders of both companies. Those conversations continue, he replied, but the priority lies elsewhere. Our strategy, at least at this point, is to spend a lot of time with the frontier models, he said, citing Anthropic, OpenAI and Google. It doesnt mean that we wouldnt consider Harvey or Legora, but prioritization is probably at the frontier labs at this point. His reasoning is that corporate customers tell Relativity they run more than one frontier model, often alongside Harvey, Legora or both, and that the frontier labs are the surest to endure. I dont think Harvey and Legora are going away anytime soon, he said. I think these are real valuable, high-performing companies. But he said Relativity expects there will be at least one frontier model on the desktop of every law firm, of every company on the planet that we serve. Brown added that has been the plan all along. That really has been our intent, to go to the frontier models, frontier labs first, he said. In 2027, Brown said, Relativity will make matter intelligence available through its MCP connector, so that an authorized user