Keeping online transactions secure is no small feat for banks. Instances of financial crime proliferate as fraudsters seek to impersonate individuals and companies in order to bypass ID verification and extort money.
Financial institutions (FIs) have reacted in response, boosting their cooperation with fintechs and looking for ways to protect their customers in the face of ever-changing fraud techniques.
However, the fight against fraud is in full swing.
According to GlobalData’s “Thematic Intelligence: Cybersecurity (2023)” report, managed security revenues – involving services that are outsourced to an external vendor or supplier – will reach $135bn by 2030, marking a CAGR of 11.4% between 2022 and 2030.
Another GlobalData survey conducted in September 2022 found out identity theft came third among top concerns shared by citizens from the UK, US, France, Germany and Poland.
AI/ML in IDV is an effective gatekeeping tool – it can catch fraud at the front end of an onboarding process and slash operation costs later on.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below formBy GlobalData
Russ Cohn, general manager of EMEA at OCR Labs, tells RBI that the challenge ahead for his business is to create digital identity solutions that are accessible and inclusive for everyone. OCR Labs provides end-to-end user ID verification solutions to companies including financial institutions.
The success of any IDV solution, Cohn says, does not just rest on AI/ML technologies; rather, any IDV software should be able to share the knowledge learnt in one part of the network across the entire client ecosystem.
Q: AI has come a long way in many sectors, including banking. How does AI/ML technology work in the context of IDV in banking?
The banking industry is trying to digitally transform itself and catch up to other sectors, like iGaming, employment screening, automotive, and retail. Other digitally-native or digitally-adept sectors were early adopters of AI/ML and learned what banking is trying to learn: customers increasingly want remote, online experiences, they want them fast, and they want them available 24/7. Thus, banking, in general, needs digital IDV, that is:
- fully automated to be feasible at the large scale of banking,
- inclusive/fair to serve a diverse customer base, and
- works anywhere to span cross-border, multi-language/country banking operations.
AI/ML is needed to deliver on these customer expectations, in particular generative AI-driven facial biometrics. Generative AI and deep neural networks are especially important in banking because the stakes are so high.
We’re therefore seeing the application of AI-driven biometrics in financial services by facially authenticating customers when they access their accounts or need to perform a higher-risk activity such as transferring money, opening a new account, resetting a password, or changing an address. Facial biometrics enable a check back to the original image captured of the person performing the action in real-time to prove that the real account owner is requesting those riskier actions.
Banks see huge value in stopping fraud at the entry point of onboarding and using biometrics for re-authentication because it reduces the need for additional staff and the cost of supporting these cases. On the other side, it also improves the customer experience by shortening approval times and reducing the need to shuttle customers between representatives or departments.
Q: OCR Labs provides IDV services to a wide range of clients. How do services offered to financial institutions differ from those provided to your other clients?
The banking industry is subject to the highest level of scrutiny and regulation regarding KYC, AML, and a host of other standards that need to be met for customer onboarding. Other sectors have not historically needed such high levels of assurance around IDV, so regulation is light-touch or non-existent. This means substandard, poor-performing solutions have been used in these sectors, as fraud was not a big focus.
With an ever-increasing amount of remote services businesses need both internally and for external customers, the need for stronger authentication and verification is now becoming more prevalent. Our IDV platform was coded from the ground up to address the specific pain points that banking institutions and their customers suffer.
We built our systems to meet and, in some cases, exceed regulatory standards across numerous jurisdictions to ensure that our banking customers have full confidence in our offer. We support IDV for over 16,000 forms of government-issued ID in nearly every territory worldwide, allowing banks to scale with one central provider.
Previous offerings relied heavily on human intervention over more advanced technologies such as AI and ML. Our proprietary deep neural network automatically extracts information from an ID using a smartphone camera and natural vision processing, assesses whether a document is real or fraudulent, captures a person’s facial biometrics, and matches their biometrics to their ID photo—no humans required.
Additionally, many banking customers have come to expect instantaneous access to services, from money transfers to loan approval. By removing the need for humans to be involved in these processes, automated IDV platforms like OCR Labs deliver on customers’ expectations for immediate response.
Q: What are banks’ most significant concerns regarding IDV, and how does OCR Labs cater to their concerns?
Banks care about catching fraud, creating a great user experience, and complying with regulations when it comes to IDV.
Catching fraud at the front end of an onboarding process helps to slash operation costs later on. If you can single out those looking to take advantage of new account openings for illicit activities, then you save a lot of time and money for yourself.
The same goes for other unsecured products, such as a credit card or loan. Once fraudsters spot a potential opening, they will take advantage of it until it’s shut down, often telling others to do the same in a highly coordinated flash attack.
The best form of protection here is to shut out bad actors at the beginning of the journey, but it’s a fine balance between stopping fraud and allowing the good customers through. If credentials are compromised through a breach elsewhere, then the technology can reverify or reauthorise customers and ensure that only trusted people can use their services. These are how OCR Labs gives banks the confidence that fraud is stopped at the front door.
User experience and onboarding have become central to banking since the rise of fintechs and neo-banks. Technology should be making lives easier here, and although friction is expected when you ask for interactions, how users experience it matters.
Generally speaking, users are happy to undergo ID checks if the technology involved is unobtrusive and intuitive to use, such as auto-filling a form for them—especially if those checks involve copying out long ID numbers. OCR Labs’ technology does just that, which makes customers’ journeys easier.
Regulatory compliance across multiple markets and geographies is always challenging, so banks seek partners of the highest calibre.
Q: Consumer privacy has become a leading concern among providers of goods and services. How does OCR Labs deal with customer privacy concerns coming from banks and other financial institutions?
Preserving user privacy is paramount in our increasingly digital world. New regulations across various jurisdictions and ever-evolving cybersecurity threats have made taking protective measures all the more urgent. That is why we have prioritised privacy-by-design principles, embedding them into our platform.
OCR Labs’ facial biometrics approach solves security and privacy concerns using a proprietary approach called Facial Tokenization, which works similarly to credit card tokenisation. This method involves generating thousands of data fragments or “tokens” from the original biometric data and storing them in separate, secure databases to be used for identification or authentication purposes.
If one of the tokens is lost or compromised, the system can use one of the remaining tokens (from among the thousands available) to identify or authenticate the person. In this manner, the individual tokens can be used to represent the biometric data without revealing the actual private information about a particular face.
This tokenisation approach has never before been applied to facial biometrics, making us the first platform in the IDV space to solve the problem at scale.
Q: Focusing on identity fraud, how is OCR Labs coping with ever-changing fraud tactics and techniques?
The challenge is to create digital identity solutions that are accessible and inclusive for everyone while at the same time staying ahead of fraudsters by using even better technology than they do.
Synthetic media, perhaps more commonly known as “deepfakes,” uses AI to create convincing image, audio, and video hoaxes. The term describes both the technology and the resulting bogus content. Deepfakes often transform existing source content where one person is swapped for another.
To combat the use of such content in fraud scenarios, we have developed our proprietary Deepake Defender identity-proofing system. OCR Labs has trained this technology on a large dataset of AI-generated synthetic faces, and this technique allows the system to improve its accuracy in identifying—and locking out—bad actors attempting to gain access using deepfakes.
Another key part of the success of any IDV solution is the ability to share learnings, which is becoming imperative in the fight against synthetic media and deepfakes used in high-volume flash attacks. Using advanced deep neural networks and shared fraud hubs, we can apply knowledge learned in one part of the network across the entire ecosystem of customers.
Q: How do you see AI/ML in banking evolve in the upcoming years?
As we move into the future, generative AI will provide better user experiences, safer identity management, and wider access and inclusivity for more people who historically couldn’t access banking systems due to a wide range of biases. In turn, regulatory agencies will begin implementing more stringent policies to prevent algorithmic bias to ensure everyone benefits from this new tech in an egalitarian way.
To prepare for these inevitable changes, AI-based IDV companies must encode inclusivity into the algorithms that power their products. Through our Zero Bias AI™ technology, we are pioneering the use of generative AI to protect against discrimination on the basis of ethnicity, age and gender. This helps to ensure that everyone, everywhere, gets an equitable onboarding and authentication experience.
Furthermore, we have anchored this philosophy into our software development process by adopting a Zero Bias Code of Ethics that includes a commitment to building technology that strives to achieve maximum inclusion and fairness. These principles are followed by all members of the development team, from data scientists to programmers to other stakeholders.