Longeye raised $20 million to expand AI evidence review used by police, prosecutors and public defenders, as Ontario drafts criminal justice AI rules.
Longeye, a San Francisco company that sells an investigative workspace for digital evidence, raised a $20 million Series A led by N47, with participation from Andreessen Horowitz’s American Dynamism practice. The Business Wire announcement is dated October 21, 2026. The same day, Legora closed a $550 million Series E. Together those two equity rounds put the week of October 19 to 25 at $570 million on the Legal Tech Circle funding tracker, once Longeye is counted.
What makes the Longeye round worth a closer look is not the size. It is who the customers are. Police departments, sheriff’s offices, prosecutors and public defenders are already using the product on live cases, while rulemakers are still writing the rules for AI in criminal work.
What Longeye Sells
Longeye describes itself as an AI powered investigative workspace built for the justice system. Agencies use it to search and analyze evidence already collected in a case, including jail calls, warrant returns, phone extractions, video, financial records and document sets. N47’s investment note says the company offers Investigate and Monitor products aimed at detectives, prosecutors, public defenders and the wider justice ecosystem.
The company says more than 50 justice agencies now use the platform. Across those agencies, Longeye reports more than 2,300 investigations, more than 15 million pages of documents, more than 10 million images, more than 32,000 hours of audio and video, and more than a billion rows of financial and phone records. It equates that workload to more than 200,000 investigator hours. The casework includes at least 230 homicide investigations and at least 160 sexual violence cases, according to the same release.
Longeye was founded in 2025. Co founders Guillaume Delépine, the CEO, and Danielle Man both came out of the N47 portfolio, per N47. The company says it meets the FBI’s CJIS Security Policy, that agencies keep ownership of their data, and that it has filed patents on a citation engine that links findings back to the original message, document or recording. It also runs a grant program that gives free access to eligible public defender offices and defense nonprofits.
Why Both Sides of the Case Matter
Most legal tech funding still lands on tools for law firms and in house teams. Longeye is selling into a different buyer: local, state and federal justice agencies that already hold the evidence. That includes the defense. The public defender grant is a deliberate signal that the company wants the product on both sides of a case, not only with investigators and prosecutors.
Verification is the pitch buyers will hear. Longeye says every substantive finding should be checkable against the source file. N47 partner T.J. Rylander put it this way in the Business Wire release: investigators told the firm Longeye was “getting them through evidence that would otherwise have sat untouched, in a way they could check and defend.” Oklahoma City Police Department homicide detective Lauren Cunningham said the tool helped detectives find “the few minutes or few pages that matter most” in jail calls, videos, phone downloads, warrant returns and financial records.
None of that is an independent audit. The usage numbers, the CJIS claim and the citation patents all come from the company and its lead investor. Buyers should still ask for proof of data retention, audit logs, how outputs are reconstructed for court, and what happens when an opposing party challenges an AI assisted finding.
The Rules Are Still Being Written
On October 26, Law Times reported that the Law Commission of Ontario released a final report on AI in criminal justice with 29 recommendations. The report is described as the first in Canada to examine how AI systems fit into criminal work. Recommendations include a governance framework for AI used by police, Crowns, courts and other institutions, mandatory public disclosure of AI system use, required impact assessments, continuous auditing, and either amendments to Ontario’s Enhancing Digital Security and Trust Act, 2024, or new complementary legislation.
The commission also called for a joint federal provincial Criminal AI Evidence Rules Committee, more Legal Aid Ontario funding for defense counsel facing AI evidence, and an AI Centre of Excellence at Legal Aid Ontario. Report authors Ryan Fritsch and Nye Thomas wrote that AI can improve investigations and proceedings, and that systems such as predictive policing, facial recognition and generative AI have proven risks to human rights, procedure, privacy and access to justice. They also noted that no Canadian government has yet enacted laws governing AI use by police or criminal courts.
Ontario’s report does not regulate Longeye. It does show how quickly the policy conversation is moving into the same evidence rooms Longeye sells into. Add Longeye to the live equity totals and 2026 year to date funding on the funding tracker is $2.96 billion across 57 rounds. Longeye did not disclose a valuation.
Common Questions
Who led Longeye’s Series A?
N47 led the $20 million round. Andreessen Horowitz’s American Dynamism practice participated, according to Business Wire and N47.
Who uses Longeye?
The company says more than 50 justice agencies use it, including police, prosecutors and public defenders across local, state and federal agencies in the United States.
Does Longeye work for the defense as well as investigators?
Yes, according to the company. It offers a grant program with free access for eligible public defender offices and defense nonprofits.
How does this change the 2026 funding picture?
With Longeye counted, year to date equity funding is $2.96 billion across 57 rounds on the Legal Tech Circle tracker. The week of October 19 to 25 becomes two rounds totaling $570 million.
What to Watch
Watch how often AI assisted evidence review shows up in discovery disputes and court challenges, and whether US jurisdictions follow Ontario’s lead on public disclosure of criminal justice AI systems. For buyers, the open questions are still the practical ones: can every finding be traced to a source file, who owns the data, and what happens when the other side asks for the audit trail.
Real. The $20 million Series A, the lead investor, the customer base Longeye claims, and Ontario’s 29 recommendation report are all on the record. The product claims about speed and case outcomes still rest on company and investor materials.