Planning for open enrollment has already begun in earnest. Every year brings new challenges in meeting timelines, adapting to changes across clinical and healthcare environments, and helping enrollees access support & guidance so they get the best healthcare for their needs. Perhaps most urgent are cutting costs and maintaining regulatory compliance. Can AI translation help with these areas? Read our blog to get answers to your questions about AI- powered translations, how to use them, and the benefits of including AI for translation in your open enrollment health insurance planning this year.

Can I use AI to translate for open enrollment 2026?

Yes, you can use AI translation for your materials — but not by itself. This is largely because of the Affordable Care Act (ACA) Section 1557.

Section 1557 addresses AI usage and language access, requiring covered entities to take “reasonable steps” to provide “meaningful access” to people who speak limited English. Meeting this requirement must involve cost-free language assistance services, which include qualified translated documents. If machine translation is used for these types of documents, a qualified human linguist must review the finalized translations.

• Documents where “accuracy is essential”
• Documents that are “critical to the rights, benefits, or meaningful access” for LEP individuals
• Open enrollment documents
• EOC/ANOC
• Summary of benefits
• Enrollment forms
• Taglines and notices about free language services, which must be provided in the top 15 languages spoken by LEP individuals in the relevant state
• More

ACA Section 1557 doesn’t address AI for translation specifically by name yet, but does refer to its output: automated translation. It says that automated translation solutions (including machine translation and AI services) can be used. However, they’re permissible only as a part of a multi-step process. A qualified human linguist must review all finalized translations. Additionally, the ACA encourages using qualified translations and interpreters to ensure accuracy and reliability of information shared with LEP individuals.

Currently, the Centers for Medicare & Medicaid Services (CMS) has not addressed AI-powered translation services specifically, leaving Section 1557 of the ACA to provide guidance. CMS has addressed translation and language access in 42 CFR 422.2267. Here, it requires organizations offering Medicare Advantage plans to translate required materials into non-English languages spoken by at least 5% of the population in the plan service area.

Does AI translation pose risks when used for open enrollment?

Beyond the blatant compliance violations, AI translation poses risks when you don’t have localization governance from language solutions integrators (like Lionbridge). Here are two key risks that AI translation without human linguist experts brings for open enrollment materials:

• Inconsistent quality for rare and indigenous languages: Machine and AI translation have improved at astounding rates, but not consistently for all languages. Rare and indigenous languages are still typically unsupported by even the highest-quality Large Language Models (LLMs). Using AI translation for these languages without adequate human review may leave speakers of these languages with low-quality, inaccurate materials.

• Lack of Context and Industry Familiarity: AI models can predict language, but it’s hard to do so for ACA-compliant open enrollment materials without deep familiarity with insurance industry terminology and regulations. Models can be trained on this information, but it’s unlikely they can possess the full, deep comprehension of this material that human experts can.

What are the benefits of using AI translation in your language planning?

Using a hybrid translation approach with Machine Translation (MT), AI models, qualified human linguists, and expert project managers can help you reach your goals compliantly with lower costs and quicker turnaround times.

• Compliance: Lionbridge’s team has decades of experience assisting insurance companies with ISO (18587) certified human and AI workflows. We’ve helped hundreds of customers working on open enrollment health insurance translation projects, and we understand what’s required to avoid ACA Section 1557 violations.

• Faster turnaround times: With LLMs and MT engines trained by linguistic experts, you can get stronger first drafts of materials. These materials require less review time from human reviewers and help you produce your multilingual materials exponentially faster so you can meet every internal and external open enrollment deadline.

• Cut costs significantly: Integrating MT and AI models into the translation process notably reduces required human labor and review. This reduces your overall language costs.