No. Asylum claim assessment is the most fact-intensive analysis in immigration law and depends on the declaration, country conditions, corroborating evidence, and the specific asylum office or immigration judge handling the case. The system prompt forbids the bot from opining on whether nexus to a protected ground will be found, whether the persecution rises to the legal standard, or whether credibility will be sustained. Those are advice questions that belong in the paid consult with the attorney who will draft the declaration.

Yes. Defensive cases are in removal proceedings before an immigration judge, and the bot recognises the signals - NTA language, EOIR portal references, master calendar or merits hearing dates. When those signals appear, the bot captures the hearing date, the alien number, and the court location, then triggers a priority callback through your webhook. Defensive cases are time-bound and the standard intake script is the wrong shape for them.

Carefully. If a user describes a family member detained by ICE, the bot acknowledges the situation, shares the ICE Detainee Locator (locator.ice.gov) so family can confirm the facility, references the Office of Refugee Resettlement family reunification line for unaccompanied-minor cases (1-800-203-7001), and triggers a same-day callback through your webhook. The bot does not attempt to coach the family through bond strategy itself. That routing keeps emergencies out of the standard calendar.

Yes, at a high level. The exceptions to the one-year filing deadline are changed circumstances (conditions in the country of origin have materially changed) and extraordinary circumstances (the applicant faced circumstances that prevented timely filing). The bot explains the categories but does not opine on whether specific facts meet the exception standard. That analysis is fact-intensive and belongs in the consult, especially because a misjudged exception argument can sink an otherwise strong claim.

Yes. GPT-4o and Claude both handle Spanish, French, Arabic, Mandarin, Portuguese, Haitian Creole, and Russian natively, which covers most asylum populations in the US. The bot detects the user's language from the first message and responds in the same language. For markets with Pashto, Dari, Tigrinya, or Ukrainian speakers, the same multilingual capability extends naturally. Most asylum firms run a single multilingual bot rather than separate per-language bots.

Conversation logs live in your WordPress database under your control. Asylum intake involves persecution details, family information, and country-of-origin facts that are extremely sensitive. Most asylum firms purge full logs on a strict schedule (often within 7 days) and forward only summary emails to the case management system, often after redacting names. API calls use your own OpenAI, Anthropic, Google, or OpenRouter key, so prompts do not pass through any SleekAI server.

Yes. Withholding of Removal and Convention Against Torture protection are alternative forms of relief that often run alongside an asylum application. The bot can describe the higher 'more likely than not' standard for withholding and the CAT framework for torture claims, while explaining that these are alternative protections, not substitutes for asylum. The attorney decides in the consult which forms of relief to pursue and how to structure the I-589.

At a high level. The bot can reference State Department country-conditions reports and your firm's published country-of-origin pages, and can explain that country-conditions evidence is a core part of the asylum case. It does not predict whether a specific country's conditions support a specific claim, because that analysis depends on the persecution narrative and the nexus question. The bot routes country-conditions analysis to the consult with the right intake form attached.