Generic AI was supposed to flatten legal software into a prompt box. Sandstone’s $30 million Series A says the opposite: in-house legal teams still need software that knows the business context behind the document, not just a model that can write convincing clauses.
The round, led by Lightspeed Venture Partners and announced Tuesday (June 9), will fund Sandstone’s enterprise expansion and hiring across legal and technical roles, according to PYMNTS. That use of proceeds matters. Sandstone isn’t pitching legal AI as a cheaper text generator. It’s pitching a system built around context, intake, triage, drafting, review, and knowledge retrieval.
Sandstone’s $30 Million Series A Puts Legal AI Back Inside Enterprise Software
Sandstone’s argument cuts against a popular AI assumption: if models get smarter, application software gets less important. The company says legal work doesn’t fit that story.
“This is why ‘SaaS is dead, just use an agent’ gets legal exactly backwards,” Sandstone said. “The hard part was never generating the words and replacing them in Word. The hard part is binding them to context — and doing it right.”
That is the core of the raise. Sandstone is betting that legal AI value sits below the surface, in the messy link between counterparties, stakeholders, matters, obligations, contracts, and history. Artificial Lawyer described the company’s platform as “connecting counterparties, stakeholders, matters, obligations, contracts, and history into one working surface,” with AI then applied across legal workflows, according to Artificial Lawyer.
XOOMAR analysis: this is not a feature race in the usual sense. Sandstone’s challenge is proving that its AI can become part of the daily operating layer for legal departments, where context changes the answer. A generic answer that “looks right” is exactly the failure mode Sandstone is trying to attack.
The Funding Math Shows Investors Are Paying for Context, Not Just Models
The known numbers are sharp:
- Series A: $30 million
- Lead investor: Lightspeed Venture Partners
- Announcement date: Tuesday (June 9)
- Use of funds: expand the enterprise offering and hire people with both legal and technical expertise
- Earlier round: $10 million Seed round in January, according to Artificial Lawyer
- Recent company claim: revenue increased by over 40x in the past 90 days, according to Artificial Lawyer
- Named customers: Wayfair, Grindr, Mercury, Cox Media, and ElevenLabs, according to Artificial Lawyer
PYMNTS also places Sandstone inside a wider legal AI funding push. Harvey raised $200 million in March for legal infrastructure for law firms and in-house teams, in a round valuing the company at $11 billion. PYMNTS also cited its earlier reporting that funding to legal technology startups surpassed $2.4 billion in the first nine months of 2025.
That doesn’t prove Sandstone wins. It does show investors are still assigning serious value to legal workflows where documents, decisions, and institutional memory collide.
The contrast is useful:
- Before: Legal AI was often judged by whether it could draft, summarize, or search.
- After: Sandstone is arguing the defensible layer is knowing which draft, summary, or answer fits a company’s specific relationship and history.
That distinction echoes a broader software tension we’ve tracked in AI workflow automation tools that can burn cash if buyers don’t compare first. Automation only compounds value when it is aimed at the right workflow. Otherwise, it just moves work into a new interface.
Lawyers and Technologists Are the Same Hiring Problem Here
Sandstone says the new capital will help it recruit “across every function” for people who understand both the profession and the technology.
“Our customers run some of the most complex legal organizations in the world. We’re investing in the people and infrastructure to onboard them fast and support them obsessively,” the company said. “We’re recruiting across every function for people who understand the profession and the technology — because that combination is the whole point.”
That line is doing more work than a normal hiring announcement. Sandstone is saying legal AI can’t be built by technologists guessing at lawyer behavior, or lawyers manually supervising a thin AI wrapper.
The company’s critique of generic AI is blunt. A generic tool may answer the same question as a senior lawyer in a way that appears right, but still fail because it lacks the firm’s business context. Sandstone calls that “confident and context-blind.”
XOOMAR analysis: this is why the hiring plan is part of the product strategy. If Sandstone wants to sell an enterprise legal platform, it needs people who can translate legal work into software primitives without stripping out the context that makes the work legally useful.
Co-founder and COO Jarryd Strydom described the operating reality to TechCrunch, quoted in Nairavoice: legal work arrives through “different intake channels,” including Slack messages, emails, Jira, and AI helps “route and triage that work appropriately,” after which teams can build custom workflows for drafting, reviewing, or legal analysis.
That is a different frame from a standalone legal chatbot. The product has to sit where the work begins.
Sandstone Is Fighting the “One Expensive Path for Every Task” Problem
Sandstone’s cost argument is also specific. PYMNTS reports the company’s view that legal work often defaults to expensive frontier models or outside firms even when the task does not require that level of horsepower.
“Not every task needs the senior law firm or Opus model. Most legal tools can’t tell the difference. We can, because we know what the task actually is,” Sandstone said.
That is the cleanest explanation of the company’s positioning. Sandstone wants to route work by task type and context, not by habit.
The strongest version of the thesis looks like this:
| Legal work assumption | Sandstone’s counterargument |
|---|---|
| A stronger model solves the workflow | The model still needs business and relationship context |
| Every task can use the same level of intelligence | Different legal tasks require different levels of support |
| Document automation is enough | Intake, triage, history, obligations, and stakeholders all matter |
| AI replaces the software layer | AI needs the software layer to reach the right context |
This also explains why Sandstone is not positioned exactly like Harvey in the source material. PYMNTS says Harvey’s AI-powered products target workflows in contract analysis, due diligence, compliance, and litigation. Sandstone’s own language focuses heavily on in-house legal context and the relationship behind the work.
The Competitive Pressure Comes From Both Startups and Frontier Labs
Sandstone is not entering a quiet category. PYMNTS points to Harvey’s March raise and Anthropic’s expansion of Claude tools for the legal industry.
Anthropic said legal professionals had become the “most engaged Claude Cowork users” of any knowledge-work function after the release of its first legal plugin. The company then expanded with “a much larger set of tools.”
That creates a clear tension for Sandstone. Frontier AI labs can improve model capability quickly. Larger legal AI companies can raise more capital. Sandstone’s defense, based on the sources, is specialization: the “legal-specific context” and workflow layer for in-house teams.
The same pressure is hitting SaaS more broadly. As we wrote in $500M Lovable run rate puts SaaS vendors on notice, AI-native tools are forcing buyers to ask whether old software categories still deserve their budget. Sandstone’s answer is not to abandon SaaS. It’s to make the application layer smarter, more contextual, and more embedded in the actual work.
Buyers Should Test Sandstone’s Context Claim, Not Just Its Demo
For legal teams evaluating AI tools in 2026, the useful question is not “Can it draft?” Sandstone itself says generating words has become the easier part.
The sharper buyer questions are:
- Context: Does the system understand the relationship behind the work, or only the document in front of it?
- Workflow: Can it handle intake across the channels legal teams actually use, including email, messaging, and business tools?
- Task matching: Can it distinguish between work that needs a heavier model, a lighter workflow, or human review?
- Integration: Can it connect to existing legal and business systems rather than becoming another isolated tool?
- Repeat use: Does it become part of daily legal operations, or only a one-off drafting assistant?
Those questions come directly from the gap Sandstone is trying to exploit. The company’s raise will look justified if enterprise customers use it as infrastructure for legal work, not as a novelty interface for AI-generated text.
The next evidence to watch is concrete: more named enterprise deployments, proof that Sandstone can keep expanding beyond early customers, and signs that its “context layer” actually reduces the default reliance on the most expensive path for routine legal work. If those signals appear, the $30 million Series A will mark more than another AI funding round. It will show that legal AI’s winning layer may be the system that knows when not to use the biggest hammer.
The Bottom Line
- Sandstone’s $30 million raise shows investor demand for legal AI built around enterprise workflows.
- The company is betting that legal teams need context-aware systems, not just document generation.
- The funding will support expansion and hiring across legal and technical roles.
Originally published on XOOMAR. For more news and analysis, visit XOOMAR.
