Federal language access policy has been unsettled for two years. It’s about to get more concrete, not less. One bill working through Congress would restore and codify federal language access standards agency-wide. A separate, narrower bill covering telehealth has already been signed into law, with a hard compliance deadline attached. Neither depends on the other, but both point in the same direction: for agencies, health systems, and the vendors that support them, language access is moving from a policy preference back toward an enforceable requirement, and the operational readiness question is the same regardless of which bill you’re tracking.
The One That’s Already Law: The SPEAK Act
Before there was a bill to restore language access government-wide, a narrower one focused specifically on telehealth already crossed the finish line. The Supporting Patient Education and Knowledge Act (SPEAK Act) was first introduced by Rep. Michelle Steel in 2023, then reintroduced by Rep. Monica De La Cruz (R-TX) in March 2025. It became law on February 3, 2026, passed as Section 6213 of the Consolidated Appropriations Act, 2026, rather than as standalone legislation.
The SPEAK Act is narrow by design, but it’s not symbolic. It gives the Department of Health and Human Services one year, until February 3, 2027, to publish guidance on language access best practices in telehealth, developed in consultation with language service companies, interpreter and translator associations, and LEP patient advocates. That guidance must address five specific areas:
- Integrating qualified medical interpreters into telehealth encounters as part of the clinical workflow, not as a bolt-on step
- Delivering appointment access details and platform instructions to LEP patients in their preferred language before the encounter begins
- Making patient portals meaningfully accessible in multiple languages, covering results, messages, and care instructions, not just an English-first interface with partial translation
- Supporting reliable multi-person video technology that connects clinician, patient, and interpreter simultaneously with consistent audio and video quality
- Delivering multilingual materials around the visit itself, including appointment reminders and prescription information
For health systems and their language partners, that’s not a future policy conversation. It’s a one-year clock that started the day the President signed the bill, and it directly shapes what “compliant telehealth interpretation” is going to mean in practice starting in 2027.
The One Still Moving: The Language Access for All Act of 2026
The broader effort is still in motion. In January 2026, Reps. Grace Meng, Judy Chu, Dan Goldman, and Juan Vargas introduced the Language Access for All Act of 2026 (H.R. 7223) in the House. In July, Sens. Andy Kim (D-NJ) and Mazie Hirono (D-HI) introduced a companion bill (S. 4985) in the Senate. Together, the bills would formalize and strengthen federal language access standards across every agency, not just healthcare, and create a coordinated accountability framework.
The bill has not become law. It remains in committee review, referred to the House Committee on Oversight and Government Reform, with no further action recorded since introduction. Its path forward is genuinely uncertain. But its intent is clear, and worth understanding now.
At its core, the Language Access for All Act would codify Executive Order 13166, the 25-year-old framework that directed federal agencies to improve access to their programs for individuals with limited English proficiency (LEP) and update it for how agencies operate today. Specifically, the bill would require agencies to:
- Maintain formal language access plans: within a year of enactment, each agency would need to develop, publish, and regularly update a plan describing how it ensures individuals with LEP can meaningfully access its programs and services, not just informally accommodate them on request.
- Use qualified linguists for critical communications: the bill sets professional standards for interpreters and translators, moving agencies away from relying on untrained bilingual staff or family members, particularly for high-stakes interactions where a mistranslation carries real consequences.
- Set explicit standards for AI-assisted translation and interpretation: rather than banning AI outright or leaving its use unregulated, the legislation treats translation technology as part of the toolkit while requiring that AI-assisted content involving critical communications be reviewed by qualified human translators before it reaches the public.
- Create formal reporting and accountability mechanisms: a public complaint system would let individuals report when they were denied meaningful access due to language barriers, with the Attorney General publishing an annual report broken out by agency, language, and program.
- Extend access into digital and clinical channels: the bill’s provisions reach beyond printed materials into telehealth interpretation and multilingual patient portals, the same ground the SPEAK Act already covers for HHS specifically, but applied across every federal agency.
Taken together, these provisions would take a framework that already existed in practice for a quarter-century and place it in statute, where it can’t be undone by a single executive order.
How We Got Here
- 2000: President Clinton signs Executive Order 13166, directing federal agencies to improve language access for LEP individuals in federally conducted and federally funded programs. This becomes the operating framework for the next 25 years, under both Democratic and Republican administrations.
- 2022–2023: The Department of Justice and individual agencies strengthen implementation of EO 13166 through updated guidance and refreshed agency-level language access plans.
- Late 2023: Rep. Michelle Steel first introduces the SPEAK Act, focused specifically on telehealth language access.
- March 2025: President Trump signs Executive Order 14224, designating English as the official language of the United States and revoking EO 13166. The same month, Rep. Monica De La Cruz reintroduces the SPEAK Act.
- July 2025: DOJ guidance follows, encouraging agencies to scale back “non-essential” multilingual services and lean on AI translation tools as a cost-saving alternative wherever legally and operationally feasible.
- January 2026: House sponsors introduce the Language Access for All Act (H.R. 7223), framing it as a restoration of a standard that had been in place for a generation.
- February 2026: The SPEAK Act is signed into law as part of the Consolidated Appropriations Act, 2026, starting HHS’s one-year clock to publish telehealth language access guidance.
- July 2026: Senate sponsors Andy Kim and Mazie Hirono introduce the companion bill (S. 4985), arguing that no single administration should be able to unilaterally cut off access to federal services based on the language someone speaks.
Why the Durability Distinction Matters
The strategic difference between an executive order and a statute is the whole reason the Language Access for All Act exists. An order can be revoked by the next president with a signature, which is exactly what happened to EO 13166 after 25 years of continuous operation across administrations of both parties. A law passed by Congress can’t be undone the same way; reversing it requires another act of Congress, not a change in who occupies the White House.
The SPEAK Act is proof of that logic in practice. Whatever happens to EO 13166’s broader legacy, the SPEAK Act’s telehealth requirements are now statute, sitting outside any single administration’s reach. That’s the model the Language Access for All Act is trying to extend agency-wide: less about creating requirements from scratch, more about restoring and legally protecting a framework many agencies and vendors were already operating under, while modernizing it for how agencies use AI, telehealth, and digital portals today.
Why This Matters for Language Access Programs, Regardless of Outcome
The Language Access for All Act’s fate in Congress is genuinely uncertain, and that uncertainty is itself the planning problem agencies and vendors need to solve for. Waiting to see whether it passes before preparing is a bet that demand for demonstrable language access capability disappears if the bill fails. The SPEAK Act’s fixed February 2027 deadline says that bet is already lost for at least one part of the landscape.
The underlying need hasn’t moved either way: more than 25 million people in the U.S. have limited English proficiency, and they still need to interact with federal, state, and local government programs and the health systems serving them. Whether that need is met under a revived statutory mandate, HHS guidance tied to a hard deadline, or state-level requirements that increasingly diverge from federal policy, the operational capabilities required to meet it are the same. Bid materials and program documentation should already be able to demonstrate:
- Qualified linguistic and interpretation resources across the languages an agency’s or health system’s constituents actually speak, not just the most common ones
- Reporting and complaint-tracking capabilities that can produce the kind of disaggregated data the Language Access for All Act would require agencies to publish
- Accessibility across formats and channels, including alternate formats like large print, braille, and audio alongside digital and telehealth access
- Documented quality assurance processes, especially for any AI-assisted translation or interpretation, with a clear record of human review
- Secure, compliant technology that can stand up to scrutiny on data handling, retention, and access controls
- Program-level governance, not just point solutions purchased ad hoc as needs arise
Agencies and vendors that treat these as baseline requirements now, independent of what happens with either bill, will be in a stronger position than those waiting for legislative certainty that may not arrive on any predictable timeline.
What the Market Is Already Signaling
Set the legislative uncertainty aside for a moment, and the direction of the market is clear. Four shifts are showing up regardless of how either bill fares:
- Specialized AI, not general-purpose AI: generic machine translation tools perform unevenly across languages, and the gap is largest in less commonly spoken languages where agencies and health systems have the least internal capacity to catch errors. The market is moving toward AI trained and fine-tuned for regulated, high-stakes content, not one-size-fits-all engines.
- Integrated workflows over point tools: organizations juggling separate systems for translation requests, interpreter scheduling, quality review, and complaint tracking are discovering how much risk lives in the gaps between those systems. Consolidated platforms that connect translation, interpretation, and reporting are becoming the expectation, not a nice-to-have.
- Technology-enabled interpreters, not AI replacing interpreters: both bills’ language reflects this. The Language Access for All Act prohibits fully replacing human linguists with AI, and the SPEAK Act’s focus areas assume a qualified human interpreter is present in the encounter, with technology supporting that connection rather than substituting for it.
- Demonstrable governance over stated policy: it’s no longer enough to say an agency, health system, or vendor has a language access policy. Bid evaluators, auditors, and now federal statute are asking for documentation: who reviewed what, when, under what standard, and what happens when something goes wrong.
What to Prepare Now
Whether or not the Language Access for All Act becomes law, the SPEAK Act’s February 2027 deadline gives every health system and language partner a concrete forcing function to act on today:
- Inventory current AI use in translation and interpretation workflows, including tools being used informally by staff without a formal review process
- Classify communications by risk tier, so that AI-assisted workflows are reserved for lower-stakes content while critical communications route to qualified human review
- Document the human review process in writing: who reviews AI output, what standard they apply, and how corrections get tracked
- Map current telehealth and portal capabilities against the SPEAK Act’s five focus areas now, rather than waiting for HHS guidance to arrive in February 2027
- Audit reporting capabilities against what the Language Access for All Act would require: complaint tracking, disaggregation by language and program, and public reporting readiness
- Review vendor security and data practices, particularly for any AI tools handling constituent or patient information
How BIG Language Solutions Can Help
BIG Language Solutions helps government agencies, health systems, and their partners build language access programs that hold up under scrutiny, whatever happens with pending legislation. Our secure, compliant infrastructure, including HITRUST, ISO 27001, and SOC 2 Type II certifications, pairs with a risk-based human-in-the-loop review model so agencies and health systems always know when a qualified linguist has reviewed critical content before it reaches the public or a patient.
LanguageVault® gives agencies and health systems transparent reporting and access controls that align with accountability requirements like those in the proposed Language Access for All Act, while our network of qualified interpreters and translators covers 300+ languages for the complex, high-stakes communications that AI alone shouldn’t handle. LanguageExpress™ supports the high-volume, time-sensitive translation demands of large-scale public programs and telehealth operations, and our consulting team works directly with agencies and health systems to build or update the formal language access plans and telehealth readiness this legislation, enacted or proposed, requires.
Ready to get ahead of the February 2027 deadline or build a language access program that’s ready for whatever comes next in federal policy? Contact us.
