Up for discussion during our latest Wine Webinar Experience this time were the correlated areas of Machine Learning (ML) and Artificial Learning (AI). We heard, specifically, from IT professionals working in the Finance industry, gleaning valuable insight into the uptake, use cases and future value to the Finance industry of both ML and AI. As an introduction, when we talk about machine learning, think about the technology that drives things like chatbots, predictive text and even how Netflix suggests certain films or shows to you. Machine learning is probably best described as a subset of artificial intelligence whereby computers are able to learn without being programmed. It’s fair to say that the terms ML and AI are commonly used in a reciprocal fashion.
Automating the backend
AI is already in use within the finance industry, albeit in a background fashion. Organisations will be using AI quite heavily for automatic document filling, things like online applications for example. And the more it gets deployed, the more people are able to be freed up to work on development which in turn can make things quicker for the end user on the customer side. For example, when a customer is filling out a mortgage application you will undoubtedly have AI scanning documents, checking them, verifying them and generally looking for anomalies in the customer’s applications. And AI/ML is also appearing in the likes of operational management and security tools – most of the big vendors are really beginning to push machine learning in these areas. And whilst some of those in the finance sector might not perceive any ‘direct’ contact with AI/ML, they will do so tangentially via cloud adoption. Much of the security built-in by cloud vendors is a real value add because many of the perimeter defence and software defined controls are all predicated upon – the big data and the associated machine learning that goes into that to enable the identification of indicators of compromise etc.
Many IT professionals working in the finance industry are no doubt familiar with the challenges faced from the board or executive teams who seem to believe that everyone else is using AI/ML so why aren’t they? This is mainly due to the fact that many of them are using it in the background but it is not ‘obvious’ to those not working directly with it. A much more obvious use case could be the whole area of Anti-Money Laundering (AML), for example. offers scope as a very worthy use case. Many firms are still relying on manual labour when it comes to AML compliance, which isn’t ideal from a performance or cost point of view. Indeed, with the adoption of the Anti-Money Laundering Act of 2020, our friends across the pond face new challenges in AML enforcement and regulation – surely a compelling reason to make AI/ML raise its game here?
AI and cloud security
Those with Microsoft expertise who might be working on data protection projects (GDPR, for example), will certainly be familiar with AI/ML. Microsoft’s trainable classifiers allow you to basically feed machines with examples of documents, allowing that machine to learn and to subsequently go off and label said documents based on company policy. This is a particularly compelling use case for financial services given the likes of Financial Conduct Authority (FCA) GDPR regulations that provide constant challenges for board or C suite. We briefly touched upon the area of cloud security. The big cloud service providers (AWS, Azure, Google etc) all possess massive pools of data. They are therefore perfectly placed to interrogate that data in the face of a cyberattack to see if there are any similarities to previous incursions – is it familiar to any of their signals?
Data gathered by service providers enables them to build AI learning into an alert which can then be used by a SOC. So, one of your engineers may be using just such an alert thinking that it’s merely a part of the software when in fact, AI underpins it.
One area that could do with improvement to facilitate the adoption of AI/ML within financial services is to open the lines of communication between various organisations. Within this sector, it’s fair to say that businesses don’t talk to each other very well. In many cases, firms are heavily reliant on vendors who do lots of work at scale with other FS organisations, to be able to identify issues or problems that they’ve seen in common scenarios. So when it comes to any potential indicators of compromise, they have enough data (from their work in this sector) to protect firms. By relying almost exclusively on vendors with their scalability, firms get what they need without having to talk to any other peer companies in their industry. So perhaps it’s time for a sector with a reputation for closed-door mentality to help drive the adoption of AI/ML to the benefit of all within financial services.
Continue reading…
- Creating a Cyber Security Culture with former Arsenal F.C IT Director, Christelle Heikkila
- Demand for ZTNA continues its upward trajectory in 2022
- What does “cyber resilience” mean to Legal IT?
- Where are you on the machine learning and artificial intelligence roadmap?
- Voices of NetMotion: reflecting on 2021