During a keynote speech at a conference at Seattle University School of Law in early March, Rachel Rossi, director of the Justice Department’s Office for Access to Justice, proclaimed that “Access to justice means language justice.” She was emphasizing the moral and legal burden U.S. courts have to enable everyone with a legal challenge to access the courts in a language they’re proficient in.

Is the United States court system meeting this burden?

To be sure, courts are making improvements. A recent survey of state courts by the National Center for State Courts (NCSC) found that 93% of respondents had improved language access over the last five years, especially in the area of expanding awareness of the need for adequate services. The American Bar Association has established comprehensive standards for language access. And Rossi’s own office has completed a large-scale translation project to make more court information available to litigants in their own language.

Despite this progress, however, leaders from the legal profession, the Department of Justice, and advocates for equitable court access agree there’s more work to be done. While recognizing courts’ progress, NCSC research has identified significant areas where improvement is still needed, including providing interpreters in civil cases, providing interpretation services outside of the courtroom, and training interpreters on how to support victims in domestic violence and sexual assault cases.

The Brennan Center for Justice lays the challenge out more starkly, based on its study of the availability of interpretation services in 35 states. Of the courts studied, 46% didn’t require interpreters to be provided in all civil cases, 80% failed to guarantee the court would pay for interpretation services, and 37% didn’t require the use of credentialed interpreters when they were available.

As long as language access remains uneven in state courts, millions of Americans will remain underserved by the justice system. Defendants will be less capable of defending their cases. Civil litigants will be hampered in their pursuit of justice. And perhaps most importantly, countless injustices will continue unchecked because those harmed—from victims of financial crimes to those living through domestic violence and sexual abuse—simply can’t access the system that’s designed to save them.

The challenge is especially severe for asylum seekers, who must navigate a complex legal process in order to have their applications granted. Just like other Americans, language access is a civil right for asylum seekers, guaranteed under Title IV of the Civil Rights Act. But as those arriving in the United States come from increasingly diverse language backgrounds, the materials and interpreters aren’t always available to provide them the services they deserve.

Across state courts, the obstacles to providing adequate language access is nearly universal: there aren’t enough interpreters to go around, and the cost of interpretation and translation services can be prohibitive for court budgets. To overcome these obstacles, some leaders are looking to generative AI, with its capacity for contextually accurate, natural language interpretation, as a possible solution. Conceivably, as the technology matures it could lead to dramatic enhancements in language access.

In the meantime, however, Rossi is right: without language access, there is no access to the U.S. system of justice. As courts struggle to provide those services, they’re also failing to protect the civil rights of the 30 million Americans who lack English proficiency.