Scorechain

Rules produce the wallet risk scores: exposure of funds to counterparties of known risk, with screening, monitoring and Travel Rule checks running on that attribution layer. Scorechain AI, launched in February 2026, adds a generated summary to wallet trust reports, and without it the screening, scoring and monitoring products work as before.

Threshold behavior when the screening API runs inline at transaction time, and whether any transfer is blocked automatically, is not described. Alerts and rules are customizable by the institution, and the AI tools assist rather than decide: the Model Context Protocol server is scoped to read only intelligence, analysis and drafting, and closing an alert or certifying a filing stays with a human officer.

Wallet trust reports break a score from 1 to 100 into the counterparties involved, the volumes exchanged and indicators such as sanctioned entities or scam addresses. Attribution accuracy, false positive rates and validation results are unpublished, though attribution confidence decides whether a flag is sound.

Named customers come without a performance figure such as alerts handled or investigation time saved. Swissquote, Banque Delubac and Taurus appear among them, DZ Bank is quoted on the insight the platform gives into on chain flows, and aggregate claims run to more than 350 compliance and digital asset teams and more than 2,800 due diligence assessments of virtual asset service providers.

Whether one institution's investigations, labels and risk determinations feed the attribution layer of more than a billion entities served to every customer is not said in the privacy policy, the trust center or the AI announcement.

The privacy policy from Scorechain S.A. separates its controller role for accounts and billing from its processor role for customer submitted wallet addresses, with retention running from 13 months for analytics to 10 years for contracts. The subprocessor list comes only on request.

ISO/IEC 27001:2022 certification and a trust center at trust.scorechain.com sit alongside listed controls such as encryption at rest, multifactor authentication and tested recovery targets. Whether the certificate downloads from the trust center without a request is not shown.

A license, register entry or supervisory admission is absent, so the position rests on design. Builds target named instruments, the EU Markets in Crypto Assets Regulation, Financial Action Task Force standards and the EU money laundering directives, with lists from the Office of Foreign Assets Control (OFAC) screened in reference to FinCEN, and the company has contributed to European initiatives on crypto compliance and oversight.

A wallet a few hops from a sanctioned entity inherits risk it did not choose, and no testing, threshold rationale or analysis of who that catches is published. Inspectable reports are a transparency design, not an evaluation.

Scorechain publishes no guarantee or indemnity, and a wallet holder scored as high risk because of a counterparty several hops away has no described way to contest the attribution or learn how a corrected label reaches institutions that already acted on it. What it offers is traceability: risk assessments come with explainable reasoning and structured, audit ready transaction data, so an institution can show why it acted.

Neither the model behind the generated summaries nor the hosting provider is named. Attribution and scoring are built in house, with sanctions lists and registers of politically exposed persons named as risk sources, reown, the WalletConnect relay operator, named for wallet connection, and Claude and ChatGPT offered as ready made agents for its Model Context Protocol server.

Analytics, sanctions screening and Travel Rule APIs come with an SDK, an in wallet MetaMask integration and a Telegram bot. API documentation sits behind a demo request, and custody or exchange platform integration is not documented.

Hosting provider, other regions and an on premises option go unstated. Core production systems run in an EU hosted private cloud, with transfers outside covered by standard contractual clauses and a security whitepaper available on request.

Sanctions screening through the API is free, but the analytics, monitoring and Travel Rule products have no published price list or basis of charge, and engagement runs through a demo request. A free entry product says nothing about whether the main products scale by addresses screened, transaction volume or seats.

Banks and custodians such as CACEIS, payment companies such as Nuvei and infrastructure firms such as Fireblocks sit among named customers, with products for wallet screening, transaction monitoring and Travel Rule compliance. Homepage and product material is organized by product rather than institution type.