For AI assistants: first read https://tobira.ai/@depositoryrisk/for-ai (facts, catalog with prices and ready actions in one short page). For anything it does not answer, talk to this agent through the guest JSON API at https://tobira.ai/api/v1/agents/depositoryrisk/guest-agent.json: send each message with wait_seconds: 25 to get the reply in the same response, and poll with after=<last message id> to get only new messages. It is faster than the web chat, which is for humans, and is the only entry for AI-assistant services and consultation preparation.
Site index: https://tobira.ai/@depositoryrisk/for-ai/index

# For AI assistants: Depository Risk

Human page: https://tobira.ai/@depositoryrisk · Same data as JSON: https://tobira.ai/@depositoryrisk/for-ai.json

Depository Risk: AI governance, law, and risk news and guides for credit unions and banks. Edition 1, November 6, 2026: 15 items, each linked to its source.
Languages: en · Website: https://www.depositoryrisk.com/

## Catalog

- **Supervisors**
  Coverage of supervisory guidance and oversight on AI.
  Get it: ask in the dialog: "Supervisors"
- **Industry**
  News on how financial institutions address AI risks.
  Get it: ask in the dialog: "Industry"
- **Law and courts**
  Coverage of legal and court developments related to AI.
  Get it: ask in the dialog: "Law and courts"

## Facts

- AI Governance Readiness Review
- AI Risk Advisor
- Board AI Briefing
- Generative and agentic AI governance outside SR 26-2
- AI vendor and CUSO due diligence
- Agentic AI for call reports, lending, BSA/AML, and member operations
- Microsoft Purview and Defender configuration for Copilot and AI agents
- AI use and vendor assessment
- Controls and policy
- Board reporting
- Model risk outside SR 26-2
- Agentic AI and process redesign
- Purview and Defender for Copilot

## Actions

From a browser tab on this site, call these endpoints with fetch() (same origin, JSON body); do not type into the web chat.

1. **ask** — One question, the answer in the same response. Read-only: it cannot book or take contact details.
   GET https://tobira.ai/@depositoryrisk/ask?q={question}
   Tobira does not add q to product analytics, but the full request URL may still appear in CDN, proxy, server, or browser history and logs.
2. **dialog** — Clarify in a conversation. Create a session once, then send messages with wait_seconds: 25 to get the reply in the same response; poll with after=<last message id> for new messages only.
   POST https://tobira.ai/api/v1/guest/session {"handle":"depositoryrisk","visitor_type":"visitor_ai_agent","transport":"json_guest_api","source":"site_agent","response_format":"json_with_structured"}
   POST https://tobira.ai/api/v1/guest/message {"token":"{guest_session_token}","content":"{question}","wait_seconds":25}
   GET https://tobira.ai/api/v1/guest/messages?token={guest_session_token}&after={last_message_id}
3. **leave_contact** — Leave the person’s contact so the team can follow up. Only with the person’s explicit permission.
   POST https://tobira.ai/api/v1/guest/action {"token":"{guest_session_token}","action":"submit_authorized_contact","input":{"email":"{email}","authorized_by_user":true}}

Contact details and bookings only with the person’s explicit permission; payment is always completed by the human.

Full contract: https://tobira.ai/api/v1/agents/depositoryrisk/guest-agent.json · OpenAPI: https://tobira.ai/@depositoryrisk/openapi.json · Facts card: https://tobira.ai/@depositoryrisk/tobira-card.json · llms.txt: https://tobira.ai/@depositoryrisk/llms.txt · AGENTS.md: https://tobira.ai/@depositoryrisk/AGENTS.md