Building the Best Blockade Against Fraud

The fraud dilemma those working in financial services are constantly facing is unlikely to end, seeing fraud surge 24% during the COVID-19 pandemic alone. Financial institutions (FIs) are still struggling to find the right solution that can balance making real use of their data via analytics and collaboration, whilst remaining completely compliant with regulators to avoid major fines.

In the first half of 2022, global financial institutions were fined over £186,208,915 ($224 million USD) for anti-money laundering (AML) compliance breaches alone. In the bid to avoid these fines and remain completely compliant, many FIs may fear making full use of all possible data analytics available. According to FICO, 56% of British consumers would leave their bank if they were found to have been involved in a money laundering scandal, most likely as a consequence of passive monitoring. Finding balance between this ‘lose-lose’ scenario is tough, and no one wants to be the next victim of a major fine.

‍Embracing technology in the fight

It's promising to see a mass embrace from FIs of technological innovation in the fight against fraud. Most notably, huge waves have been made in harnessing artificial intelligence (AI) technology to analyze large amounts of publicly available transactional data, filter out false alerts and spot patterns of potentially criminal activity that are too complex to be picked up by the human eye. Yet, still, despite a significantly increased focus on implementing emerging technologies to enhance fraud, AML and know-your-customer (KYC) models, financial criminals are only getting more sophisticated. According to the World Economic Forum, financial crime continues to account for 2-5% of global GDP.

This all comes back to the stringent regulatory backdrop having an impact on the extent to which FIs can make real use of required data to monitor and implement intelligent decisions. The financial services sphere is still seeing only a fraction of the entire picture of their customers’ activity, making it incredibly difficult to spot patterns and criminal behaviour. Wouldn’t it be ideal if financial institutions could meet all regulatory and compliance obligations, whilst continuing to increase data utility through both intra and inter-company collaboration – in turn improving all services needed to crack down on fraud, money laundering and illicit finance?

Confidentiality by design

Although the embrace of AI to enhance many fraud-detection models has been revolutionary, it can go further. Machine-learning based models are still largely being trained on synthetic data, or not making the most of all internal and external data available due to fears of breaching intensely monitored regulations.

Embracing confidential computing and its spotlight feature, the trusted execution environment (TEE), a secure area of a central processing unit (CPU), might be the answer. Within a single company, confidential machine learning within an TEE could enable FIs to use real customer data when training their models, improving the ability to identify fraud whilst also remaining on the right side of all privacy regulations. This would also reduce significant costs, not needing to purchase external expensive datasets — win-win! Consequently, critical data analytics based on machine learning models can be embraced to the fullest, without fears of fines.

On top of this, using a secure TEE, that inherently encrypts all data throughout its lifecycle – including the critical processing stage – provides provable trust via attestation. With this, multiple service providers will feel at ease sharing customer data across datasets. Not only now at ease, but also without any risk of breaking regulations, and knowing that details will never be visible to other providers that may give them a competitive advantage. Another win-win. This remote attestation feature in confidential computing is one of the critical cogs in fostering a new era of digital trust, enabling the requisite revolution of data collaboration to truly win the fight against financial crime.

Ditching the data dilemma

So, if you’re experiencing issues between facing the potentially damaging consequences of a major fine but want to make full use of all data available to you, confidential computing provides the building blocks to overcome this.

The time is now to upgrade your financial crime monitoring models all round, harnessing the power of confidential data analytics and multi-party data collaboration.

Don’t know where to start? Get in touch.