Case study · Scam prevention

Reasonable steps you can prove: preparing an Australian bank for the Scams Prevention Framework

A customer-owned Australian bank had fraud controls built for stolen cards and taken-over accounts, not for customers persuaded to send money themselves. The Scams Prevention Framework turns scam prevention into a legal duty with a regulator watching. The goal was a platform that finds scams earlier, acts on intelligence quickly and can show, payment by payment, why it did what it did.

Case studyRetail banking · AustraliaSumanta Mukhopadhyay9 min read
#security#ai#data
CASE STUDY · CASE STUDY · SCAM PREVENTION18 Nov 2026 · 15:50 GST
At a glance
SectorRetail banking · Australia
My roleFractional CTO and AI architect
EngagementReadiness assessment, architecture and delivery leadership
ScopeScam intelligence, real-time payment risk, disruption, reporting, complaints and reimbursement, evidence
What changed
  • Scam signals from payments, devices, customers and partners in one case view
  • Higher-risk payments slowed with targeted friction instead of blanket holds
  • Intelligence turned into disruption and regulator reports as tracked work
  • Every intervention recorded with the signals and rule or model behind it
  • Complaints and reimbursement handled from the same evidence
Payment event streamReal-time risk scoringScam entity graphIntelligence case storeCustomer friction serviceLLM report triageDecision evidence storeOpenTelemetry tail sampling

Anonymised composite. Client names, identifying details and confidential figures are withheld; public facts are cited in the references.

The situation

A mid-sized, customer-owned Australian bank had sound fraud controls for the problems of a decade ago: stolen card numbers, compromised passwords, account takeover. Its losses had shifted to a different pattern. Customers were being talked into sending money themselves, through investment, impersonation and romance scams, and the bank's systems saw a properly authenticated customer making a real-time payment. Every control said the payment was legitimate, because technically it was.

The Scams Prevention Framework changes what a bank is expected to do about that. Passed in February 2025, it sets overarching principles for designated sectors: governance, prevent, detect, report, disrupt and respond1. When a bank has actionable scam intelligence, meaning reasonable grounds to suspect a scam connected with its service, it must report it to the ACCC and take reasonable steps to disrupt it, with a safe harbour for disruption while it investigates, for up to 28 days2.

Why now. Banking, telecommunications and digital platforms have been designated as the first sectors, and the framework is moving from principles to detailed sector codes4. Treasury released draft rules and codes, including a common code and a banking code, for consultation in May 20263. The obligations are close enough that the platform work had to start now, not after the final text.

What had to be true

  • Scams treated as their own problem, not as a subset of unauthorised fraud.
  • One view of a scam across payments, devices, customer reports and intelligence from other organisations.
  • Friction that matches risk: most payments untouched, the risky few slowed in ways that actually break the scammer's script.
  • Intelligence acted on in hours, with reporting and disruption tracked as work, not emails.
  • Evidence for every intervention, strong enough for a complaint, an external dispute or the regulator.

Architecture

Signals
Paymentsreal-time, card, billDevice and sessionremote access, new deviceCustomer reportsapp, phone, branchExternal intelligencebanks, telcos, platforms
Detection
Real-time risk scoringrules plus modelScam entity graphpayees, mules, numbers, sitesLLM report triagestructured, human checked
Action
Friction servicewarn, confirm, hold, callDisruptionblock payee, restrict muleReportingregulator and sharing
Response
Intelligence casesowner and clockComplaints and reimbursementinternal disputesCustomer supportafter a scam
Assurance
Decision evidence storeimmutableTracesfull for flagged paymentsControl testingscheduled
Signals feed one detection layer; every action it takes leaves evidence behind.

The data flow, end to end

  1. InitiateA customer starts a payment; payment and session events go on the stream
  2. ScoreRules, model and graph lookups score the payment in the payment path
  3. InterveneHigher risk gets targeted friction: a specific warning, confirmation or hold
  4. LearnOutcomes and customer reports flow back as intelligence
  5. Disrupt and reportActionable intelligence opens a case to block, restrict and report
  6. EvidenceThe decision, signals and actions are stored for complaints and review
Detection in the payment path, disruption and reporting as tracked cases, and evidence at every step.

The hard part: proving reasonable steps, case by case

A framework built on reasonable steps is judged after the event. A customer complains weeks later; the question is what the bank knew at the moment of the payment, and what it did. Fraud systems usually keep a score and an outcome. That is not enough. The design stored an immutable decision record for every payment that was scored above the lowest threshold: the signals present, the rule or model version, the friction shown and the customer's response.

Traces were the second layer. Keeping a full trace for every payment would be expensive and mostly useless. Tail sampling in the OpenTelemetry Collector makes the keep-or-drop decision after all spans in a trace are complete5, so the team kept every trace for a flagged, held or interrupted payment and a small sample of the rest.

Diagram of a tail sampling processor containing a trace with a root span and two child spans, deciding whether to keep the trace only once all spans are done
Decide what to keep once the whole trace is complete: every flagged payment, a sample of the rest. Source: OpenTelemetry blog, Reese Lee · CC BY 4.0

Traces cannot be the only record. A scam reported a week later was never flagged at the time, so its trace may be gone. The decision record is the evidence; traces are the detail when you still have them.

The entity graph followed the approach in the fraud ring knowledge graph, using entity resolution to connect payees, mule accounts, phone numbers and websites across reports. The LLM triage step turned free-text customer reports into structured intelligence, with a person confirming anything that would trigger disruption; it was evaluated on real, de-identified reports first, as in evaluation sets that predict production.

How it ran

Risk and compliance
  1. Map principles to controls
  2. Define actionable intelligence
  3. Board and regulator reporting
Engineering and data
  1. Payment event stream
  2. Scoring, graph and friction
  3. Evidence store and traces
Operations
  1. Scam case handling
  2. Customer friction scripts
  3. Complaints and reimbursement
Compliance defined the thresholds, engineering built the paths, operations ran the cases.
Phase 1Assess
  • Scam loss patterns
  • Gaps against principles
  • Draft code review
Phase 2See
  • Event stream
  • Entity graph
  • Case store
Phase 3Act
  • Risk scoring
  • Targeted friction
  • Disruption and reports
Phase 4Prove
  • Evidence store
  • Control testing
  • Track final codes
See scams first, act on them second, and make the proof a product, not an afterthought.

Key decisions and trade-offs

DecisionAlternativeWhy
Separate scam models from fraud modelsExtend the existing fraud modelThe customer is genuine; the signals and the right response are different
Targeted friction, such as a scam-specific questionGeneric warnings on every paymentCustomers learn to click past generic warnings
Intelligence as cases with owners and clocksAlerts in a shared inboxReporting and disruption must be timely and provable
Person confirms LLM-extracted intelligenceAutomatic action on model outputWrongly blocking a payee harms a real customer
Decision records for every scored paymentRely on logs and tracesLogs rotate; complaints arrive weeks later
Build to the principles, track the codesWait for final codesThe principles are stable; the detail can be tuned

What I'd do again, and differently

Again: start with the evidence record and the case model, because they shape everything else, and treat the regulator's definition of actionable intelligence as an engineering requirement with a threshold and an owner. Differently: bring operations and the contact centre in on friction design from the first week. The best interventions came from the people who talk to scam victims, not from the risk model. Related patterns are in idempotent payment pipelines and deepfake-resistant onboarding.

If you run a regulated Australian business, my work with Australian companies explains how I approach it, or analyse your challenge and bring the result to a call.

References
  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
FAQ

Common questions

What is the Scams Prevention Framework?

An Australian law passed in February 2025 that makes designated businesses, starting with banks, telcos and digital platforms, take reasonable steps to prevent, detect, report, disrupt and respond to scams, under principles overseen by the ACCC and sector codes enforced by sector regulators.

What is actionable scam intelligence?

Information that gives an entity reasonable grounds to suspect that a communication, transaction or other activity connected with its service is a scam. It can come from customers, the entity's own analysis or other organisations, and triggers duties to report it and take reasonable steps to disrupt the activity.

Does a bank have to stop every scam?

No. The framework is built around reasonable steps. That is why the evidence matters: a bank needs to show what it knew, when, and what it did about it, for each case.

Keep reading

Related

Open to advisory, fractional CTO & select builds

Let's talk about what you're building.

The fastest route is a free 30-minute call. Prefer to write? Send a message and I'll reply personally within two business days.

Start a conversation

How would you like to reach me?

Book a 30-min callFastest. Pick any open slot on my calendar.→ Write in GmailOpens a pre-filled draft in your browser.→ Use my email appOutlook, Apple Mail or your default client.→

I reply personally within two business days. Time zones from the Americas and Europe to the Middle East and Asia-Pacific.