Lilshira
AI tools not reliable for immigration law advice, report says
A Vals AI report warns that AI chatbots are unreliable for immigration law advice, especially for temporary protected status and DACA cases. No new rule is introduced, but the findings affect anyone using AI for immigration information.
In this story
At a glance
- What changes
- No new legal rule introduced but AI advice reliability highlighted
- Timing
- Report released November 6 2026
- Who is affected
- Anyone using AI tools for immigration information
The details
- Announced by
- Vals AI
Report Overview
Vals AI evaluated popular AI models on immigration‑law questions and compared answers to expert rubrics. The assessment covered models such as Anthropic’s Claude, OpenAI’s ChatGPT and Google’s Gemini.
Key Findings
Even the best‑performing model gave incomplete answers in most scenarios. Models struggled most with questions about temporary or fragile status, including TPS and DACA. AI tools were also poor at reflecting policy changes from 2025 and 2026.
- Best model still incomplete in majority of cases
- Least accurate on TPS and DACA queries
- Recent policy updates often missed
Implications for Users
The report advises against relying on AI for specific immigration advice. Users are encouraged to seek professional legal counsel and verify information with official sources such as USCIS or reputable immigration‑law organizations.
- Do not use AI as sole source for legal decisions
- Cross‑check AI answers with official agency guidance
What this means for travellers
Trip planning
- Existing travel bookings are unaffected by the report
- No changes to entry or transit requirements are introduced
Cost and documents
- No new fees or document requirements are created
- Paperwork for immigration applications remains unchanged
What you need to do
- Watch for updates from official immigration agencies
- Verify any AI‑generated advice with a qualified lawyer before acting
Who this affects
-
People seeking immigration advice via AI
The report shows AI answers are often incomplete or incorrect
What to do Should verify information with professional counsel
-
Temporary Protected Status (TPS) applicants
AI models performed poorly on TPS questions
What to do Must rely on official guidance for TPS applications
-
DACA recipients
AI models struggled with DACA‑related queries
What to do Should consult qualified immigration attorneys for DACA matters
Frequently asked questions
Can I trust AI chatbots for immigration advice?
No, the Vals AI report found that even top‑performing models give incomplete answers, especially on temporary status issues. Professional legal counsel is recommended.
Which AI models performed best in the immigration test?
Anthropic’s Claude, the most expensive version tested, was the best‑performing model but still provided incomplete answers in most scenarios.
Do AI tools know about immigration policy changes from 2025‑2026?
The report indicates AI models often missed recent policy updates, providing less than a third of correct information on 2025‑2026 changes.
Are TPS and DACA questions especially risky for AI answers?
Yes, the study highlighted that AI models were least likely to give complete and correct answers for temporary or fragile status topics like TPS and DACA.
What should I do if I receive AI‑generated immigration advice?
Treat AI output as informational only; verify any advice with an immigration lawyer or official agency resources before taking any action.
