Skip to main content
LLM Squared

For community banks and credit unions.

Check AI use and the evidence still missing.

Answer eight scoping questions. Check the records that fit. Keep the summary on this device until you choose to share.

Check AI use, then the matching records

Preparation aid, not an exam result. A Yes identifies activity, not accepted evidence. Record status is a separate choice.

Work stays on this page. No answers are counted, saved automatically or sent. Attach a summary only if you choose.

Save and reopen this record check

JSON retains these fixed choices. Reopening requires a new summary and does not restore review acceptance.

Does the institution use AI in products, services, operations or internal support?
Has the institution identified the AI systems and uses it currently has?
Does AI face customers or support or influence decisions?
Does the institution rely on vendors or external platforms for AI?
Has the institution checked products for embedded AI and asked vendors about it?
Does the institution use generative AI or language models?
Does the institution separate AI uses by risk, impact or review needs?
Does sensitive or confidential information pass through AI tools or services?

Record checklist

The fix order puts open questions first, then baseline governance and inventory records, then other requests. This is a local preparation order.

On hand means a record exists. It does not mean anyone inspected or accepted the record. Confirm exclusions with the reviewer.

Prepare a tier for one AI use

Choose each factor rating from the optional CSBS worksheet. The highest factor sets the preliminary tier, not an institution approval.

Open the official tiering worksheet. Suggested controls stack by tier. Your institution assigns and approves its final tier.

Source: Core Examiner Guide, pages 3, 4 and 7. Checked October 16, 2026.

Read all 19 possible requests. The matching checklist is a local preparation mapping, not an agency scoring rule.

Earlier saved checks and eight-record workflow

Preparation aid, not an exam result

Choose Current record only when an up-to-date record exists. This self-check does not inspect sources or judge whether evidence is sufficient.

Save and reopen a check on this device

Download the JSON file after preparing a summary. Reopen that file to restore the eight selections and prepare a fresh summary.

Files stay on your device. This check has no automatic storage or upload. Keep client and account data out of the file.

Where does AI support your institution?
Which AI tools and uses have owners?
Does AI face customers or influence decisions?
Which AI capabilities come from vendors?
Where is AI embedded in vendor products?
Which generative AI tools are in use?
How does your institution separate uses by risk?
What sensitive information passes through AI?
Compare an earlier saved check

Open an earlier readiness JSON file to compare with your current selections. Both checks stay on this device.

Prepare a current summary, then choose an earlier saved check.

Changes in reported records do not show source changes, closed findings or institution approval.

Owner, reviewer and institution decision fields stay blank. Keep the original check files to reopen selections later.