We've had a busy first half of 2026 at Inscribe. We've introduced new ways for risk teams to investigate applications faster, create custom risk signals, tune fraud detection to their unique risk profile, and bring Inscribe directly into existing decision workflows. Here's a look at what's new.

Risk teams often need a second opinion during a review. That might mean understanding why a document was flagged, checking whether bank statement activity matches policy, reconciling figures by hand, or looking across multiple submissions for signs of coordinated fraud. The Assistant is built for those moments.
You can ask it to explain a fraud signal in plain language, cross-reference all documents uploaded by a customer, check for gambling or crypto transactions in a bank statement, reconcile figures that would otherwise require manual calculations, or surface trends across recent submissions that could indicate coordinated fraud activity.
Every answer is grounded in the documents and signals available in your Inscribe workspace, helping reviewers get context without leaving their workflow. The Assistant is available throughout the Inscribe web app, including from the top right of any screen, the Customer view, the Document view, or directly from a fraud signal. Wherever you are in a review, it is within reach.
Learn more about the Assistant.

Off-the-shelf detection can only tell you what a platform was trained to flag. But every institution has its own risk policies, thresholds, and fraud patterns. A transaction that looks suspicious to one lender may be expected behavior for another. Custom transaction insights let your team define the bank statement activity you want to detect.
You can build rules based on transaction category, merchant, payment processor, payment method, amount, or timeframe. You can set balance-based thresholds, track average daily balance, flag volatility against a rolling average, or trigger signals based on raw transaction count or percentage of total activity. Before a signal goes live, you can backtest it against your existing applicant population to see how many customers would have been flagged.
Once your signals are live, Inscribe runs the analysis automatically. Analysts do not need to manually cross-reference transactions or remember institution-specific criteria. Instead of relying on a generic risk score, your team can evaluate bank statements through the lens of your own policies.
Learn more about custom transaction insights.

Fraud patterns change over time, and document fraud detection should be able to change with them. The redesigned Decision Engine gives risk teams more flexibility to tune Inscribe around the fraud you're seeing, rather than relying on a fixed set of rules built for a generic risk profile.
You can now tune detector severities, customize decision logic, manage blocklists, and incorporate custom transaction and document insights into your review process. Acceptance criteria can also be tailored to align with your institution's risk strategy. The result is fraud detection that becomes more precise over time — as your team learns which signals matter most, Inscribe can be tuned to help reduce false positives, catch more true fraud, and reflect how risk shows up in your queue.
For teams using Taktile to manage decision workflows, Inscribe's agentic document fraud detection is now available directly inside the Taktile platform. This makes it easier to flag high-risk documents earlier, route legitimate applications through faster, and reduce manual review volume without sacrificing accuracy.
Learn more about the integration here.

I'll admit some bias on this one. I love and only use dark mode, so it felt worth calling out! It's a small addition compared to some of the releases above, but one that makes long review sessions a little easier on the eyes.
Inscribe now supports a consistent light and dark mode experience across the application, automatically adapting to your browser settings. Current and future experiences are designed to work seamlessly across both themes.

If you're an existing customer and would like a walkthrough of any of these releases, reach out to your Customer Success manager. If interested in learning more about Inscribe, book time with our team!
Stephanie Spangler is the Head of Product Marketing at Inscribe, where she covers AI-powered fraud detection, document risk, and how financial institutions are adopting agentic AI. She writes on the intersection of product and practice — translating what fraud detection technology does into what it means for the risk teams using it.
Start your free trial to catch more fraud, faster.