Most of a commercial credit request is preparation: reading statements and returns, typing them into a spread, reconciling them, working the ratios, adding up the grade points, valuing the collateral, testing the policy and writing the memo. Very little of it is judgement. That makes credit a good fit for AI agents, and a risky one, because the small part that is judgement (whether to lend, on what terms, and whether an exception is acceptable) is exactly what a credit officer signs for and an examiner asks about. The design question isn’t whether agents can do the work. It is which decisions they must never take, and how that is enforced rather than promised.

In short

  • Give each agent one job on the request. Eight agents, each with a narrow role: read the documents, spread and reconcile, value the collateral, grade, test the policy, draft the memo, check what blocks, and watch the loan book.
  • Agents prepare; people decide. Anything that reaches the borrower, clears a check, justifies an exception, orders an appraisal, puts a borrower on the watchlist or decides the request is a prepared decision a person accepts, rejects or asks about.
  • Autonomy is a dial with a hard ceiling. Every action starts at “Draft”. No credit agent can go above “Ask first”, whatever its record.
  • Figures come from the engines, not the model. Every number an agent writes comes from the reader, the spread or the grade. With the model switched off, the agents still run, on rules.

Eight agents, one request

InsightCredit runs eight agents on the platform’s agent layer. Seven work a credit request in order, from intake to submission; the eighth watches the loan book. Each one does the preparation for its step and, where it finds something only a person can act on, prepares a decision.

AgentWhat it doesWhat it can prepare for a person
Intake agentReads statements, returns, bank statements, agings, rent rolls, personal financial statements and appraisals into fields, each quoted from its lineA request to the borrower for what is missing
Spreading agentSpreads three fiscal years and reconciles them with the tax returns, the agings, the bank statement and the rent rollA request to the borrower to explain the differences
Collateral agentValues the collateral from the appraisals and the aging at the bank’s advance rates, computes LTV and coverage, flags appraisals too old for new moneyAn appraisal order, for a person to place
Credit analyst agentComputes the scorecard grade from the ratios, global cash flow, collateral and guarantors, and explains what drives itA request to the borrower to explain a decline in earnings
Policy agentRuns the bank’s credit policy controls and drafts a mitigant for each exception from the case’s figuresThe drafted mitigants, for the analyst to edit and record
Memo writerDrafts the credit memo from the spread, the grade, the collateral and the exceptionsThe memo, for the analyst to accept or edit
SupervisorChecks what blocks a decision and which approval authority the request needsSubmitting the request for approval
Covenant monitorTests covenants on each borrower’s quarterly reporting, flags declining coverage, lists late packages and overdue annual reviewsWatchlist additions; opening the overdue reviews
The InsightCredit agent catalogue: eight agents across four steps, nine governed actions, each with what it does and what it prepares, and a note that with the model switched off the agents fall back to rules
The catalogue. Eight agents across four steps and nine governed actions. In the demo the model is switched off, so every agent runs on its rules.

Approval

Analysis

Spreading

Intake

Intake agent

Spreading agent

Collateral agent

Credit analyst agent

Policy agent

Memo writer

Supervisor

A person submits
another person approves

The case agents run in the order of the stages. The last step on the path is always two people.

What no agent does

The line between preparing and deciding is drawn around actions that have consequences outside the case, or that a credit officer or an examiner would expect a named person to own.

ActionWhoWhy
Send anything to the borrowerA person, from a held requestWhat the bank asks a borrower is a communication on its behalf, and hard to take back
Clear or waive a failed checkA person, with a reasonA waiver is a judgement that the evidence is good enough
Justify a policy exceptionA person, who edits and records the mitigantAccepting a request outside policy is a credit judgement
Order an appraisalA personIt commits the bank to an engagement, and appraiser independence matters
Put a borrower on the watchlist, or clear itA person, with a reasonIt changes how the relationship is managed and reported
Override a risk gradeA credit officer, with a reasonThe grade sets the policy outcome and the scrutiny
Approve or decline a requestAn officer with the authority, never the submitterThe decision is the bank’s, and it goes on the decision record

Enforced in the product, not in a prompt

These actions aren’t forbidden by telling a model not to do them. The agents have no way to do them: the functions that send, waive, justify, order, add to the watchlist and decide take a person as the actor, and the routes check that person’s role and authority. Every action an agent can prepare is also capped by a tenant setting (CREDIT_AGENTS_MAX_LEVEL) that can’t be raised above “Ask first”.

A prepared decision

When an agent finds something only a person can act on, it prepares a decision: a title, the lines of evidence, the reason, and the action that will happen if the person accepts. On the fictional Larkfield Heating & Air request, for example, the credit analyst agent prepared “Ask Larkfield Heating & Air Inc. to explain the decline in earnings”, with its evidence taken straight from the spread: revenue of $2,044,845, $1,949,915 and $1,802,738 over FY2023 to FY2025, an EBITDA margin falling from 9.6% to 6.1%, and historical coverage from 5.32x to 1.58x. (Every borrower, person and figure here is from our demo bank, Brightwater Commercial Bank, and is fictional.) The person has three answers.

Accept

Reject

Ask

Agent finds
something to act on

Prepared decision
evidence, reason, proposed action

A person answers

The action runs
e.g. a request drafted and held

Nothing happens
the rejection is recorded

A question goes on the case
the agent can run again

Recorded
who answered, when, what it did

Every answer, including a rejection, is recorded, and counts towards that action’s agreement rate.

The Agents tab on CR-2026-000003, Driftwood Industrial Supply: three prepared decisions waiting, the spreading agent's request for an explanation, the credit analyst agent's request and the memo writer's memo, each with Accept, Reject and Ask
What is waiting on the analyst. On the Driftwood request, three prepared decisions: two requests to the borrower and the drafted memo, each with Accept, Reject and Ask.

Accepting still doesn’t reach the borrower

Even an accepted request doesn’t go to the borrower. On Driftwood, the receivables on the 2025 balance sheet ($592,463) don’t agree with the aging ($436,665), a 35.7% difference against an illustrative 10% tolerance, so the check failed and the spreading agent prepared a request for the reconciliation. When the analyst accepted it, InsightCredit created a request held for review. The analyst edits the wording, which must keep the [request link] placeholder, and sends it. Two deliberate steps, because a message to a borrower is hard to take back.

The request on CR-2026-000003 created from the spreading agent's accepted decision, held for review: an editable message containing the request link placeholder, with Send and Withdraw
Held for review. The agent drafted it; a person accepted it; the same person, or a colleague, edits and sends it. Withdrawing the request stops the borrower’s link working.

The placeholder becomes the borrower’s own link, with no login; InsightCredit stores only a fingerprint of it, and shows it once. In the demo environment nothing is delivered: the person sends the message themselves. When the borrower uploads, the document is read at once and the spread and checks run again.

The same pattern holds elsewhere. On the fictional Birchwood Commerce Center request, the collateral agent found that the property was last valued on September 25, 2023 and prepared “Order a new appraisal of 1 property”. Accepting it records the order; nothing is sent to an appraiser from the demo environment, and placing the engagement remains a person’s job. On Tidewater Industrial Properties, the policy agent drafted mitigants for two referred exceptions, such as global coverage of 1.32x with the guarantors’ cash flow and their liquid assets of $2,163,000, but the request also has a block (pro forma coverage of 0.77x, below the illustrative 1.00x floor) that no mitigant can justify: it has to be cured or declined.

Autonomy, capped

Each action an agent can prepare has a level, and every action starts at the lowest.

LevelNameWhat happensAvailable to credit agents?
L1DraftThe agent prepares a draft; a person does the workYes; where every action starts in the demo
L2Ask firstThe agent prepares the decision and asks a person before it actsYes; the ceiling
L3Act and tellThe agent acts, tells the person, and lets them pull it backNo: locked
L4ActThe agent acts on its ownNo: locked

The ceiling is a hard cap, not a default. Agreement and override rates are tracked for every action: how often people accepted what was prepared, changed it or rejected it. The autonomy page shows that record and, on the demo settings, recommends a move up only after at least 90% agreement over at least 50 decisions in 30 days, and a move down when agreement falls below 70%. A recommendation changes nothing by itself, and nothing moves above the ceiling. An administrator can also hold any agent lower.

The InsightCredit autonomy page: the tenant ceiling at L2 Ask first, levels L3 and L4 locked, and each agent's actions at L1 Draft
Autonomy. The ceiling at “Ask first”, the two levels above it locked, and every action at “Draft” on the demo bank.

The model is optional; the figures are not

Each agent has a rules path that works with no language model at all. The demo runs with the model switched off (CREDIT_AGENTS_LLM=0), and every screenshot here comes from that mode. When a model is switched on, it can improve the wording of the memo and the reasoning, but its output is checked: the memo writer works from numbered facts the engines produced, and a sentence carrying a number or a name that isn’t among those facts is thrown away, leaving the engine’s wording in its place. The arithmetic (coverage, leverage, global cash flow, grade points, lendable value) is never asked of a model.

GENERATIVE The model writes the memo

Fluent, and occasionally wrong: a coverage ratio that isn’t in the spread, a guarantor’s net worth that isn’t on the statement. Nobody can tell which sentences to trust.

GROUNDED The model rewords facts the engines produced

Every number and name is checked against the numbered facts. An ungrounded sentence is discarded and the engine’s draft stands. A person still accepts the memo.

On the loan book, too

The covenant monitor applies the same rule after the loan is booked. On the demo book it ran 701 covenant tests on 257 loans and found 6 breaches, 5 borrowers with coverage still passing but falling, and 5 quarterly packages late, with 46 annual reviews due within 60 days, 13 of them overdue. It prepares watchlist additions for breaches and declining coverage, and on its last run it prepared “Open 13 overdue annual reviews”. A person adds a borrower to the watchlist and clears it, with a reason: on the demo book the 11 borrowers there (6 covenant breaches, 5 declining coverage) were added by a credit officer. Late packages go to the relationship manager to chase, not to the watchlist.

Who does what on a request

A credit request with the agents runningOwners in the tag
1
Run the agentsanalyst

The case agents run in order; each records what it did, what it found and what it prepared.

2
Answer what was preparedanalyst

Accept, reject or ask on each prepared decision. Every answer is recorded and counts towards the action’s agreement rate.

3
Do the judgement workanalyst

Resolve or waive checks with a reason, send held requests, edit and record mitigants, edit the memo.

4
Run againanalyst

When the borrower answers or a document changes, the spread and checks rerun and the agents run again on the new figures.

5
Submitanalyst

The supervisor agent says when nothing blocks and which authority is needed; the analyst submits.

6
Decidecredit officer, senior credit officer or committee

Someone other than the submitter, holding the authority the exposure and exceptions require, approves or declines with a reason.

What it means for each audience

For the chief credit officer

The boundary between automation and judgement is written into the product, not into a procedure manual. Every agent action has a level, a ceiling, an owner and a history.

When an examiner asks “who decided?”, the answer is always a named person with the authority to decide.

For the credit team lead

Analysts start each request with the reading, spreading, reconciling and drafting done, and a short list of decisions waiting.

Agreement and override rates show which agents to rely on, and which rules or settings need work.

For the board

AI shortens the path to a credit decision without moving accountability. That is the version of AI in lending that survives an examination.

Autonomy is held at “Ask first” at most; any change to that is a governance decision, not a software update.

What to be careful about

  • Agreement can be lazy. If analysts accept everything, the agreement rate rises without meaning anything. Sample accepted decisions the way credit risk review samples files.
  • Agents inherit the documents’ gaps. An agent can’t spread a year the borrower didn’t send or read a scan with no text layer. The prepared decision says what it looked at; read that.
  • The policy is yours. The controls, tolerances, advance rates and authorities the agents apply in the demo are illustrative. A credit officer edits them in Rules Studio and the bank’s settings.

Let agents do everything that is preparation. Make sure they can’t do anything that is a decision.

Narrow roles, prepared decisions, autonomy capped at asking first, and every figure from the engines.

Keep reading