Accend (YC S23)

If you’re a fintech or digital bank thinking about security - fraud algorithms and compliance aren’t your only levers. But most fintechs treat security like it's separate from the user experience. Good UX might be your best defense against security threats. It doesn't have to be that way. Look at Netflix. Their "continue watching" feature is great for users. It also flags account sharing. Or take Spotify's collaborative playlists. They boost engagement and build trust networks at the same time. I’d love to see more creative ways for digital banking to follow suit and blend security with user experience. What if payment limits changed based on how you normally use the app? What if we used social proof to build trust, not just boring verification docs? What if security features actually made the product more fun to use? Here's my prediction: The most secure digital bank won't JUST have a great compliance team. It'll be the one that makes security feel invisible through brilliant design. Any one see any examples of this happening IRL?

If you’re a bank or fintech looking to build brand trust, you should be looking at how your marketing and data security teams are interacting. At first sight, both teams seem to want the opposite. Marketing wants to use insights from customer data for targeted campaigns, and security wants to protect that data. But with all the compliance failures happening recently, customers are becoming more sensitive to data privacy issues. This is where the marketing-security partnership can become a differentiator. Companies that can demonstrate they have strong security advantages are likely to gain the competitive edge. This builds consumer trust. It also positions you as a leader in responsible data management, which builds regulator trust. The first step: getting your CMO and CISO to work together 🤝

There’s a big conflict happening in embedded finance right now. Fintechs are partnering with smaller sponsor banks to provide services. While these embedded finance partnerships drive a lot of revenue and deposits, they come with a price tag. 💸 - 75% of sponsor banks report compliance violation losses of $100k+ ❗- 80% of sponsor banks report managing and meeting compliance requirements is hard 😣 - 20% of sponsor banks cite reputational damage as a detrimental outcome of compliance failures 🤖 - And, my favorite stat, 94% are investing in new compliance tech. As the embedded finance ecosystem matures, banks are facing a critical decision: invest in robust compliance solutions and fully commit to the space... Or risk exiting the business entirely.

🚨 The Office of the Comptroller of the Currency has issued an enforcement action against Wells Fargo after finding “deficiencies” in its AML protections. Specifically, in the bank’s customer due diligence, customer identification & beneficial ownership programs. This is a MAJOR setback for Wells Fargo, financially + reputationally. Not only did their stock drop 4%, but it’s a dark cloud that will take a while to clear. The last time they got an AML consent order in 2015, they were only released from it 5 years later. The takeaway: adequate compliance is never just a tick-box exercise. Back-to-back negative regulatory action erodes the trust regulators and customers have in you. Investing in compliance and AML with the aim to EXCEED at it is a strategic business move. If you want to do this quickly without assembling a large team of analysts/BPOs, we can support you at Accend (YC S23). Feel free to DM me! Could other banks be at risk of a similar OCC order? 👀

Most compliance leaders at top fintechs tell us their outsourced AML analysts aren’t worth the cost. And it never surprises us. Here’s why. 1) Human-for-human swap doesn't reduce errors. Outsourcing merely replaces your in-house team with external analysts. The potential for human error remains unchanged, if not increased due to less familiarity with your specific processes and risk profile. 2) Quality assurance still falls on you Outsourcing companies will never care about your compliance as much as you do. The responsibility for quality control ultimately remains with your internal team, adding another layer of work instead of reducing it. 3) Scaling becomes a recurring challenge For growing companies, the need to scale compliance operations is constant. Continually hiring more outsourced analysts to meet this demand leads to rapidly escalating costs, creating a financial burden that can outweigh the output you’re getting in return. If this is something you're struggling with right now, drop me a message! I’m more than happy to help you think through this.

If you’re buying or using AI professionally (so… everyone nowadays), you need to ask vendors about explainability before you commit to them! Explainability in AI means being transparent about the way it works. In the compliance world, this is important for 2 reasons: → Explainability helps analysts to understand system outputs simply and quickly. If analysts understand how the analysis works, they can make informed decisions efficiently. → It makes it possible to understand why a mistake was made and even train the system to stop it from happening again If you’re heading up compliance, different stakeholders will need different levels of explainability. Externally, regulators will want to see that you have solid systems that are detecting what you’re expecting them to. So the explainability side there is very much around “is this AI system doing everything correctly and being able to prove and test that?” Internally, compliance and risk analysts are making decisions on the data produced by AI. So their primary focus is NOT on how an AI model has been tested or what algorithms are being used. Their focus is on, “Why do I have this result? Is this something that I could be confident in or is it something that I need to investigate further?” They need to be able to understand the 'why' rather than the 'how'. When you’re looking into AI for compliance, either to build in-house or buying an external tool, make SURE the people building it have thought carefully about explainability for different audiences.