Why We're Co-Leading DynamoFL's Series A + Our Diligence Notes on Generative AI in Banking
INVESTING IN DYNAMOFL
Artificial Intelligence (AI) is not new to the tech world, but its promises have often been larger than its delivery. Many times, we've been excited about AI's potential, only to see the reality fall a bit short.
We think this renewed market exuberance is different. Unlike previous AI hype-cycles, Generative AI (GenAI) is tangible. The experiences are good and sometimes even magical. The technology has been available to mass-market consumers since Day 1. It’s estimated that 600M+ people have already interacted with GenAI: exploring, deploying, and experimenting across myriad use cases and contexts.
This is an innovation that needs to be embraced by the enterprise. The economic value of GenAI isn't speculative; it's demonstrable. The largest financial institutions in the United States are taking the trend seriously and developing actionable implementation roadmaps.
That said, the barriers to enterprise adoption are numerous and significant. GenAI introduces brand new risk vectors and considerations for financial institutions. There are regulatory, reputational, privacy, and compliance risks that need to be addressed before meaningful penetration is attainable. To complicate matters, very few companies have built products with these considerations in mind. The world desperately needs AI infrastructure that is purpose-built for the workplace.
This is why we couldn’t be more excited to announce our Series A investment in DynamoFL, the market leader in privacy-preserved AI infrastructure. The company has quickly become the go-to partner for commercializing AI for enterprise use-cases. As one of our bank Limited Partners put it, “Dynamo can see the forest through the trees. They’re the only company that understands our challenges”. This kind of feedback is rare, and gives us confidence that Vaik, Christian, and the Dynamo team are on the path to creating a category defining company.
BROADER MARKET CONTEXT
GenAI, a cutting-edge technology with the ability to autonomously generate and manipulate data, is gradually gaining traction in the banking sector. At Canapi, we see the transformative potential of GenAI and are committed to supporting financial institutions in their exploration of its capabilities. During our diligence process for our investment in Dynamo, we spent hours with our bank limited partners and came away with new insights into the banking industry's views and strategies regarding GenAI.
In what follows, we’ll summarize some of our diligence findings, delving into banks' perspectives, their primary concerns, and the hurdles they face in adopting this innovative technology. By comprehending their outlook and investing in solutions like Dynamo, we aspire to foster a collaborative and forward-thinking approach to GenAI integration, unlocking a promising future for financial services.
A CAUTIOUS YET INTERESTED MINDSET
GenAI is a new frontier for most banks, and while they acknowledge its potential, implementation has been modest to date. This cautious approach stems from concerns around the technology's complexity, potential risks, and the desire to maintain compliance in a dynamic regulatory environment.
However, the allure of GenAI's transformative capabilities cannot be ignored. Early adopters in the banking industry have launched pilot programs to much fanfare and have reported substantial benefits, such as enhanced operational efficiency, streamlined banking operations, improved customer experience, and higher customer satisfaction levels.
With these success stories and the breakneck pace of innovation in other sectors, banks are increasingly looking at GenAI as a strategic enabler for innovation and competitiveness. As a result, we anticipate a meaningful shift in their approach over the coming years. Budgets for GenAI initiatives are projected to see a significant upswing, with expected growth ranging from 50% to over 100%within the next two years.1
GREATEST FUTURE IMPACT ANTICIPATED IN CUSTOMER SERVICE, EMPLOYEE PRODUCTIVITY, AND SECURITY
We’ve seen three use cases continually dominate conversations on GenAI’s potential impact in banking.2
1. Elevating Customer Service and Support. Leveraging GenAI can allow banks to gain deeper insights into customer behaviors, preferences, and needs, which in turn paves the way for bespoke recommendations and real-time assistance. Applications in this space can centralize knowledge bases for call center employees, making each call more effective and dramatically shortening the time it takes to train a call center worker. Directly exposing solutions to customers could allow always-on, concierge-like support for every customer, enhancing the customer experience and freeing up bank employees to manage more high-value tasks. Canapi portfolio company Posh has been an early leader in this space, helping banks leverage generative AI to uplevel their customer relationships; we anticipate many banks’ first deployments of Gen AI will center around this critical need.
2. Employee Productivity and Internal Process Optimization. The technology's ability to automate repetitive tasks and generate valuable insights empowers employees to work more efficiently and make data-driven decisions with greater precision. Morgan Stanley made headlines earlier this year when they partnered with OpenAI to build a knowledge assistant for their wealth advisors, allowing them to quickly access the banks’ vast wealth resources to advise clients more efficiently. SouthState Bank has been an outspoken early adopter of generative AI tooling for productivity, using an enterprise version of ChatGPT as an internal search engine for employees as well as for email composition, expense reporting, SARs and more.
From optimizing knowledge management systems to automating routine tasks, GenAI holds immense promise in transforming the way banking institutions operate internally, leading to increased efficiency and resource optimization; these enhancements may prove crucial at a time when bank profitability is increasingly under pressure.
3. AML, KYC, and Fraud Detection. By integrating the technology into their anti-money laundering (AML), know-your-customer (KYC), and fraud detection processes, banks can bolster their defense against financial crimes and illicit activities. The ability of GenAI to detect patterns and anomalies in vast datasets provides an added layer of protection, reducing risks and ensuring regulatory compliance.
PRIORITIZING PILOT PROJECTS TO TEST USE CASES
The adoption of GenAI is far from a one-size-fits-all strategy. The industry seems divided between a centralized and decentralized approach. Some banks have chosen a centralized model, where a dedicated team or department manages and implements GenAI initiatives across the institution, while others prefer a decentralized approach, where different teams independently explore and implement GenAI technologies as befits their business needs.
While we understand institutions will need to adopt the approach that best fits their unique structure and strategy, we see real value in aligning around a centralized strategy, at the very least as regards vendor management and data infrastructure decisions. The absence of cross-business communication could lead to redundant vendor relationships, unnecessary tech spending and a non-optimal data stack with a foundation that inhibits future innovation.
Irrespective of the approach, most banks are adopting a cautious strategy by conducting pilot projects to test the feasibility and impact of GenAI in specific use cases. They express a strong inclination to collaborate with vendors, consultants, and external partners who possess expertise and experience in GenAI implementation.
While the landscape is still relatively nascent, we expect many banks will “test the waters” with established SaaS and banking infrastructure vendors, whose trusted relationships, in-place MSAs, and large R&D budgets can deliver early value to banks for relatively low cost. This represents a major opportunity for long-standing bank vendors like the cores (JHA, FIS, FISV) and incumbent SaaS providers to expand and deepen relationships with their customers in a time of rapid change.
Over time, as they build proficiency in leveraging these new offerings and evaluate which use cases benefit most from the application of GenAI, we believe many banks will look to partner with newer ecosystem players at the foundation model and ML Ops layers to build bespoke solutions for their customers and employees.
COMPANIES CAN HELP BANKS OVERCOME CHALLENGES IMPEDING ADOPTION
Conversations with banks reveal that while the potential benefits of GenAI are enticing, they are acutely aware of significant challenges that may hinder adoption. Three standout as common concerns:
1. Data Privacy and Security Concerns: Banks handle vast amounts of sensitive customer data, and data privacy and security are of paramount importance to them. The introduction of ChatGPT and similar GenAI applications happened quickly, leaving many banks eager to experiment but unsure as to what exactly happens with their data when entered in a GenAI interface and what risks their usage might expose. The risk of data leakage is not trivial: to get the most out of GenAI solutions, models often need to be fine-tuned or enhanced with internal data. That data comes from a range of sources within the bank and needs to be thoroughly reviewed and anonymized where necessary to prevent sensitive data from leaking into external models or forums. Once a model is up and running, the bank then needs to ensure that employees are using the application safely (not exposing MNPI or PII) and that there are sufficient guardrails to prevent security issues going forward.
These are difficult technical problems that must be resolved before banks can safely benefit from the promise of GenAI. Many companies operating in the “MLOps” segment have emerged to help banks prepare their data for consumption and mitigate ongoing security weaknesses. We expect this area to be a core investment segment for banks eager to use GenAI and a logical place to start for any institution that is beginning to build out its strategy. DynamoFL is already helping fill this role for a number of large financial institutions and we anticipate robust collaboration between DynamoFL and our partner banks in the quarters to come.
2. Regulatory and Compliance Constraints: The financial sector is one of the most tightly regulated industries, making the integration of advanced technologies like generative AI particularly intricate. Regulatory frameworks governing the deployment and use of AI applications in finance are still in nascent stages, prompting a climate of cautious anticipation among banks as they wait for more explicit guidance from regulators. The prevailing ambiguity has impelled many institutions to adopt a wait-and-see strategy. In our conversations with banks, we have heard there’s very little benefit to being first to go-live with customer facing solutions; better instead to experiment with internal use cases and wait for other institutions to go-live first or for definitive guidance from Washington.
That said, there appears to be consensus among regulators that GenAI issues should be treated as high priority, evidenced by ongoing deliberations that aim to shape and define the regulatory landscape for AI. For GenAI providers, this presents an opportunity. By addressing the compliance needs of banks and proactively monitoring and being transparent about the regulatory compliance of their solutions, providers can significantly bolster their credibility. Such an approach also paves the way for forming true partnerships with banks who are eager to work with platforms that can facilitate the seamless integration and ongoing compliance of GenAI within their operations.
3. Lack of Internal GenAI Expertise: Developing and deploying GenAI solutions requires specialized skills and expertise, which many banks currently lack. As GenAI companies engage with these institutions, offering guidance and support to bridge this knowledge gap can significantly enhance chances of successful implementation. Collaborative partnerships that provide comprehensive support and training in GenAI will be highly valued by banks seeking to harness the technology's potential.
At Canapi, we are excited to continue to collaborate with banks as they venture into the world of GenAI. We are committed to helping them identify high-value use cases, overcome challenges associated with adoption and develop strategies for integrating GenAI into their operations. Most importantly, we look forward to working closely with Vaik, Christian, the entire DynamoFL team and our friends at Nexus Venture Partners in the months and years to come as we build a Generative AI-enhanced future for banks and enterprises.
1 Canapi Ventures “Generative AI in Banking” Survey of ~30 US-Based regional and community banks
2 There are several additional use cases that banks have identified and that we will explore in future posts