Block plans to make its proprietary Cash App Score available to outside lenders through Nova Credit, expanding a credit signal built from activity inside the Cash App ecosystem beyond Block’s own lending products.
The plan could give lenders a more current view of applicants whose conventional credit files are limited. It also extends the consequences of payment and account activity: behavior such as spending, saving, paycheck deposits, peer-to-peer transfers and repayment may help shape decisions about credit cards, auto loans, device financing, personal loans and even tenant screening.
Block says customers will decide whether and how their score is shared, and Nova Credit says its infrastructure can place the score in lenders’ existing underwriting workflows. The accountability test will be whether that consent remains meaningful and whether lenders can explain, validate and challenge decisions made with a frequently updated score.
The external-lender service is planned, not yet confirmed live
Block announced the Nova Credit partnership on September 1. PaymentsJournal reported the development the following day. Block’s announcement says Cash App Score has begun rolling out to customers, but its product page describes access by outside lenders in future tense and says a select group of lenders is planned for later in 2026. Public materials reviewed for this article do not identify the first participating lenders or confirm that external underwriting decisions have begun.
That distinction matters. The current customer-facing rollout and the planned lender-distribution service are separate stages. Cash App previously launched a pilot that let selected users view a score used in Cash App Borrow, see factors associated with it and receive suggestions about actions that might improve it.
Block says the score draws on millions of first-party signals and updates in near real time. Its product page also says the model uses proprietary Cash App and Afterpay data. Block reports that the technology has approved 38% more Cash App Borrow customers at the same loss rate as traditional scores. The company also projects more approvals in auto loans and credit cards at comparable loss rates.
Those are Block’s own analyses, not independently validated results in the public sources reviewed. The release does not disclose sample construction, observation periods, model thresholds, subgroup results or how it defines the comparison score and loss-rate equivalence. Lenders should not treat an aggregate approval statistic as a substitute for validation on their own applicants and products.
Block, Nova Credit and lenders will divide control
The announced structure creates at least three control points. Block builds the score from activity in its ecosystem and says it will manage customer notifications and consent. Nova Credit will distribute the score through its Cash Flow Intelligence Platform and says it brings consumer-reporting infrastructure and a Fair Credit Reporting Act compliance framework. Each lender will decide how much weight to place on the signal and what action to take.
This division can create gaps if responsibilities are not explicit. Block’s announcement says customers can set sharing preferences without providing a third-party login. It does not publicly detail, however, the consent screen, whether permission is lender-specific or purpose-specific, how withdrawal affects data already delivered, how long a lender may retain a score, or which party handles a challenge to underlying information. The absence of those details from an announcement is not evidence that controls do not exist, but it leaves important operating questions unanswered.
Tenant screening also broadens the potential impact beyond a conventional loan application. A score derived from day-to-day financial behavior could influence access to housing if a screening provider or landlord uses it. Any deployment in that setting warrants clear disclosure of the score’s purpose, the data covered, the identity of the user and the route for correcting errors.
A score alone is not an adequate denial reason
The Consumer Financial Protection Bureau has said that the Equal Credit Opportunity Act and Regulation B require creditors to give applicants specific and accurate reasons for adverse action, regardless of the technology used. The bureau’s 2022 circular says a creditor cannot satisfy that duty merely by stating that an applicant failed to reach a qualifying score or by choosing the closest reason on a standard form when it does not describe the factor actually used.
That requirement is especially important for a near-real-time model. A lender needs to preserve which model version, score, data snapshot and principal factors informed a decision at the moment it was made. A later score change cannot erase the evidence needed to explain an earlier denial, lower limit or less favorable term.
Block describes Cash App Score as transparent and says customers can see drivers and steps that may strengthen it. Customer-facing guidance is useful, but it is not automatically the same as a legally sufficient adverse-action explanation from a lender. The lender must be able to connect its actual decision to the factors it actually considered.
Controls lenders should require before deployment
Payments, fintech and credit teams evaluating the score should treat the integration as a material model and third-party-risk change rather than a simple new data field. At minimum, implementation should address:
- Consent evidence: preserve the exact disclosure, customer choice, permitted purpose, recipient and withdrawal history.
- Data lineage: document which Cash App and affiliated-product signals enter the score, their timing and the correction process.
- Independent validation: test performance, stability and calibration on the lender’s own portfolio instead of relying only on Block’s reported approval lift.
- Fair-lending monitoring: evaluate outcomes across protected groups and investigate whether payment behavior operates as an unintended proxy.
- Reason-code fidelity: verify that adverse-action notices identify the principal factors actually used by the score and the lender’s decision policy.
- Model-change controls: retain version history and monitor drift as the score reacts to changing consumer behavior and economic conditions.
- Disputes and remediation: tell consumers which organization can correct source data, score calculations and lender decisions, with handoffs tested before launch.
Alternative data can expand access for consumers whom traditional files do not describe well. It can also make routine payment behavior consequential in new settings. The value of Cash App Score will therefore depend not only on approval rates, but on whether Block, Nova Credit and participating lenders can make consent durable, decisions explainable and errors correctable as the product scales.