Financial institutions spend heavily on meeting regulations — verifying customers, monitoring transactions for money laundering, and reporting to supervisors. RegTech (regulatory technology) uses software, data and automation to do this more accurately and at lower cost. InsurTech applies technology to insurance: how policies are sold, priced and administered and how claims are paid. Both are major segments of fintech.
What you will learn
- Describe the main RegTech applications in KYC, AML, reporting and risk
- Show how better models can reduce the cost of compliance monitoring
- Explain how InsurTech changes distribution, pricing and claims
- Recognise the risks of relying on automated compliance and pricing
Why RegTech emerged
After the 2008 financial crisis, the volume of regulation grew sharply, while data and cloud computing became cheaper. Arner, Barberis and Buckley (2016) describe RegTech as the use of technology for regulatory monitoring, reporting and compliance — benefiting both firms and supervisors, who increasingly use technology themselves (sometimes called SupTech).
Main RegTech applications
| Area | What technology does |
|---|---|
| Customer onboarding (KYC) | Verifies identity documents, matches faces with photos (e-KYC), checks sanctions and politically exposed person lists |
| Anti-money laundering (AML) | Monitors transactions for suspicious patterns, prioritises alerts, supports investigations |
| Fraud prevention | Scores transactions in real time — see AI in finance |
| Regulatory reporting | Collects data from systems and produces reports in required formats |
| Regulatory change management | Tracks new rules and maps them to internal policies |
| Risk and capital calculation | Automates calculations such as capital ratios — see capital adequacy |
| Communications surveillance | Screens trader emails and messages for misconduct |
The false-positive problem
Traditional AML systems use fixed rules — for example, "flag any cash deposit above a threshold". Such rules generate huge numbers of alerts, the great majority of which turn out to be legitimate. Each alert must still be reviewed by an analyst.
Worked example: the cost of false positives
A bank’s monitoring system generates 10,000 alerts a month. About 2% (200) are genuinely suspicious; the other 9,800 are false positives. Each review takes 20 minutes (illustrative figures).
- Current review time: 10,000 × 20 ÷ 60 ≈ 3,333 analyst hours a month.
- A new model cuts false positives by 40% while still catching the same 200 genuine cases: false positives fall to 9,800 × 60% = 5,880, total alerts to 6,080.
- New review time: 6,080 × 20 ÷ 60 ≈ 2,027 hours — a saving of about 1,307 hours a month.
- The share of alerts that are genuine rises from 2.0% to 200 ÷ 6,080 = 3.3%, so analysts spend more time on real risks.
The key test is that the new model does not miss suspicious activity the old rules caught. Supervisors expect firms to validate and document any such change.
InsurTech
InsurTech changes each stage of the insurance value chain:
| Stage | Traditional approach | InsurTech approach |
|---|---|---|
| Distribution | Agents and branches | Apps, comparison websites, "embedded" insurance at the point of sale |
| Pricing | Broad rating factors | More data, including telematics and usage-based pricing |
| Policy servicing | Paper and call centres | Digital self-service |
| Claims | Manual forms and inspections | Photo-based claims, automated assessment, faster payment |
| Products | Annual policies | On-demand, micro and parametric insurance |
The fundamentals of insurance — pooling risks and pricing premiums from expected losses — remain the same; see insurance basics.
Worked example: pricing from expected loss
An insurer covers a risk with a 2% annual probability of a 500,000 loss (illustrative figures).
- Expected loss: 2% × 500,000 = 10,000 a year.
- Adding a 30% loading for expenses, risk and profit gives a premium of 10,000 × 1.30 = 13,000.
Better data improves the estimate of the probability and size of loss for each customer. A driver whose telematics data shows safe driving might qualify for a lower premium; a riskier driver might pay more.
Parametric and micro-insurance
Parametric insurance pays a fixed amount when a measurable event occurs — for example rainfall below a set level or an earthquake of a given magnitude — rather than after assessing actual losses. Payment can be fast and cheap to administer, which suits farmers and small businesses. The trade-off is basis risk: the payout may not match the customer’s actual loss. Micro-insurance offers small, low-cost policies, often through mobile channels, extending cover to low-income households.
Risks and challenges
- Model risk and over-reliance: automated compliance can miss new patterns; firms remain responsible for outcomes even when using a vendor’s tool.
- Data protection: RegTech and InsurTech process large amounts of personal data.
- Fairness in pricing: more granular pricing can make insurance unaffordable for higher-risk groups and raise discrimination concerns.
- Explainability: supervisors and customers may need to understand why an alert was raised or a claim refused.
- Third-party and concentration risk: many firms depending on the same provider — see operational risk.
- Insurance-specific regulation: insurers must meet solvency and conduct standards, reflected internationally in the IAIS Insurance Core Principles.
Common mistakes
- Treating compliance technology as a substitute for a compliance culture.
- Switching off rules to reduce alerts without proving that risk detection is maintained.
- Buying insurance through an app without checking that the insurer is licensed.
- Assuming a parametric policy will cover every type of loss.
Key terms
- RegTech — technology used to meet regulatory requirements.
- KYC — know your customer: verifying identity and assessing customer risk.
- AML — anti-money laundering controls and monitoring.
- False positive — an alert raised on activity that is legitimate.
- Parametric insurance — insurance that pays on a measured trigger rather than an assessed loss.
Frequently asked questions
Does RegTech replace compliance staff?
It automates routine work and prioritises cases, but people are still needed for judgement, investigations and governance.
What is embedded insurance?
Insurance offered as part of buying another product, such as travel cover when booking a flight or device cover when buying a phone.
Why might my insurance price change with InsurTech?
Pricing may use more data about your behaviour or circumstances. Insurers must still follow rules on fairness and the use of personal data.
Key takeaways
- RegTech automates KYC, AML monitoring, reporting and other compliance tasks.
- Better models can cut false positives and focus analysts on genuine risks, if validated.
- InsurTech changes distribution, pricing, claims and products, but premiums still rest on expected losses.
- Model risk, fairness, privacy and accountability remain the firm’s responsibility.
