Language Access for All Act: What It Means for LEP Services
A Federal Bill to Codify What an Executive Order Once Promised
Senators Andy Kim (D-NJ) and Mazie Hirono (D-HI) introduced the Language Access for All Act in July 2026, aiming to write into permanent law the federal language access obligations that Executive Order 13166 established 25 years ago. The bill comes directly in response to President Trump's March 2025 Executive Order 14224, which declared English the official language of the United States and revoked EO 13166 outright.
That revocation matters. EO 13166 was the foundation most federal agencies used to justify and fund translation and interpretation services. Without it, agencies have legal cover to cut those services entirely. The Department of Justice has already issued guidance that pushes multilingual services aside in favor of English assimilation programs.
The Language Access for All Act would reverse that. And if it passes, it would do so in a way no future executive order can undo.
What the Bill Actually Requires
The legislation goes well beyond restoring the old executive order. It creates a durable, accountable framework with real compliance teeth. Key provisions include:
| Provision | What It Does |
|---|---|
| Mandatory agency language access plans | Each agency must develop, publish, and submit plans to Congress |
| Public complaint system | LEP individuals can formally report access failures at specific agencies |
| Language access coordinator | Every agency designates a point of contact for implementation |
| AI guardrails | Automated language tools must be tested for bias and cannot replace qualified human translators or interpreters |
| Technical standards | Applies to all agency content, including digital and automated communications |
| Interagency working group | Coordinates guidance and technical assistance across the federal government |
The AI provision is particularly significant. Federal agencies are already experimenting with machine translation to cut costs. This bill would require that any AI-assisted language service comply with federal privacy rules and undergo continuous bias testing. It also explicitly prohibits using AI to replace qualified human linguists. That is a meaningful line to draw.
Who Is Actually Affected
The numbers are not abstract. Roughly 25 million people in the United States — about 8% of the population — have limited English proficiency. Spanish speakers account for the largest share, but nearly 40% of them report speaking English less than very well.
Among Asian Americans, 32% have LEP, the highest rate of any racial group. Among Native Hawaiians and Pacific Islanders, the figure is 12%.
People in these communities are not accessing optional services. They are trying to apply for Social Security, enroll in Medicare, file for FEMA disaster relief, and open small business accounts. When language access disappears, so does their ability to use services they are legally entitled to.
What This Looks Like for Pacific Islander Communities
Here is where the general industry analysis stops and the on-the-ground reality begins.
The 12% LEP figure for Native Hawaiians and Pacific Islanders understates the challenge considerably when you look at specific communities. Compact of Free Association (COFA) migrants from the Federated States of Micronesia, the Marshall Islands, and Palau have the legal right to live and work in the United States. Many live in Hawaii, Guam, Arkansas, and the Pacific Northwest. They rely on federal programs — Medicaid, SNAP, housing assistance — at significant rates.
The primary languages in these communities are not Spanish or Tagalog. They are Chuukese, Pohnpeian, Marshallese, Kosraean, and Yapese. These are not languages with large commercial translation markets. Most generalist agencies do not have qualified linguists for them. Some do not even list them as languages they support.
If the Language Access for All Act passes, federal agencies will need language access plans that cover their actual LEP populations. In states with significant Micronesian communities, that means Chuukese and Pohnpeian translation and interpretation, not just Spanish and Mandarin. Agency language access coordinators who try to fulfill the bill's requirements by contracting only with large generalist LSPs will find gaps they cannot close.
This is not a hypothetical problem. Healthcare systems in Hawaii have faced it for years. School districts in Oregon and Washington serving COFA families face it now. Federal agencies will face the same thing at scale if this legislation becomes law.
Capacity in Chuukese and Pohnpeian is thin. Agencies and their contractors need to identify qualified providers before a compliance deadline forces the issue.
What You Should Do Before This Passes
The bill has House co-sponsors from New York and California and Senate backing from Gillibrand alongside Kim and Hirono. It is a minority-party bill in the current Senate, so its path is uncertain. But language access litigation and state-level mandates are not waiting for federal legislation. Title VI of the Civil Rights Act still applies to recipients of federal funding regardless of what the executive branch does with EO 13166.
If you run language access for a federal agency, a federally funded healthcare system, or a school district, audit your current coverage now. Map your LEP populations by language, not just by broad ethnicity. Find out whether your current vendors can actually deliver qualified interpretation in the languages your community speaks.
For language service companies that subcontract rare-language work, now is the time to establish relationships with providers who have genuine Micronesian language capacity — before demand spikes and the qualified linguist pool is already committed.
The legislation may or may not pass this session. The underlying obligation to serve LEP communities is already there. Getting your language access infrastructure right is not contingent on what Congress does next.
Manages the TXLOC platform and content.
Related articles
TMS Data Security Gaps That Put Your Clients at Risk
Your Security Policy Ends Where Your TMS Begins Your organization spent months earning ISO 27001 certificatio...
MT Software Reviews Miss a Critical Language Gap
The Comparison Chart Nobody Questions Every MT software roundup follows the same script: benchmark DeepL agai...
AI Language Mixing Is Worse in Rare Languages
When AI Picks the Wrong Language A LangOps architect at Chainels recently ran a controlled experiment that ev...