By now you’ve almost certainly heard of various versions of ChatGPT or DALL-E — programs that can, proponents say, instantly generate content that’s so good, you’d swear it was created by a real person.
These technologies are classified as generative artificial intelligence, a subdomain of machine learning that focuses on generating new data such as images, text, sounds and even code. It’s like having an army of writers, artists and engineers churning out endless amounts of original content, except that it’s just a set of trained algorithms.
Quite suddenly, these technologies have transformed AI from theoretical abstraction to something very real, easy and useful. It doesn’t require too much of a leap to imagine how generative AI might change the way people use the web. We seem to be quickly moving to a future where we can ask the internet to make something for us, have a discussion with us, or solve a specific problem.
Financial services firms are no strangers to AI. As an industry, they are outpacing nearly every other in adoption of AI/machine learning technologies. Generative AI opens a new set of use cases for bank and credit union marketing: hyper-personalized content, more engaging customer experiences, better insights, and even operational efficiencies.
But there is another side financial marketers must weigh carefully. Potential legal, ethical and security risks with using generative AI in marketing applications can’t be ignored. These are risks that will need to be mitigated with the right application design, testing and rules-based guardrails.
Applying Generative AI in Banking’s Real World
Where generative artificial intelligence gets exciting is how it converges with customer relationship management (CRM). This hybrid of generative AI and CRM relies on precise consumer data to generate content that is hyper-relevant and highly engaging, helping to improve conversion, usage and loyalty performance measures.
As financial services marketers begin to explore applications of generative AI, here are four areas where these technologies can be used in practice:
• Personalized content
• Chatbots and personal assistants
• Improved customer insights
• Additional operational benefits
AI Opportunity #1) Personalized Content
Generative artificial intelligence can enable firms to create exquisitely customized marketing campaigns across channels (i.e. email, social media, TV and other video). Specific applications include:
• Marketing messages and digital ads customized at an individual level.
Generative AI can generate headlines, text and imagery tailored to an individual. For example, banks or credit unions could target a 28-year-old single male who works in tech; lives in Austin, Texas; rents an apartment and more. These data points can feed the AI to generate ad units for bundled services, with a unique combination of product positioning, messaging, and imagery, optimized for that individual. While this kind of customization has been theoretically possible in the past, generative AI makes it practical, allowing marketers to customize at scale quickly and inexpensively.
• Personalized content production at scale.
Using generative AI, financial services organizations can publish higher volumes of relevant articles, video/animation and audio content. And they can do so faster and more efficiently. For example, trust and investment sales functions at banks can create investment newsletters tailored to the exact life stage and interests of the client, including their age, risk tolerance, assets, goals, past behaviors and interests. Such newsletters might have hundreds or even thousands of permutations, each version geared to a specific customer microsegment. And while human experts would play a crucial role in supervising and approving the content generated by the AI, this technology enables mass-customization at scale.
• Customized engagement.
As consumers engage with a brand’s site or mobile app, generative AI can fuel a highly customized, optimized experience. For example, retail banks can deliver a customized site experience for prospects based on their location and branch proximity. When combined with identity management platforms that can identify visitors at a personally identifiable information level, the experience could be even further customized to interests, life stage and potential product interests.
AI Opportunity #2) Chatbots and Virtual Assistants
Generative AI-powered chatbots and virtual assistants improve customer experience by providing 24/7 support, answering frequently asked questions, and guiding customers through complex processes. These could provide:
• More efficient, higher-quality service automation.
Generative AI’s combination of natural language processing and predictive algorithms enable better conversations with financial services customers. For example, generative AI could inform automated customer service — even scanning for tone of voice and emotional state.
• Better customer onboarding experiences.
Generative AI ensures the uploading and availability of customer information and provides highly personalized suggestions and next steps. More importantly, customer experience data can be captured and fed back into the AI, improving the system’s ability to anticipate and respond to customer needs.
• Conversational guidance, service, and education.
When used with integrated CRM platforms, generative AI can comb through all the customer case notes ever generated on the platform and instantly generate relevant content. This content can be reviewed and modified by a human agent (i.e., advisor, service representative, branch personnel) before being presented to a customer, either live or through a virtual assistant or chat function. The key is that generative AI delivers a faster, more informed response than traditional methods.
- The Strategy Behind Truist’s AI-Powered Digital Assistant
- BofA Combines Marketing & Digital Banking with Powerful Results
AI Opportunity #3) Improved Customer Insights
Generative artificial intelligence can help financial services companies gain a deeper understanding of their marketing by analyzing large amounts of customer data, contact history, and campaign performance.
This information can be used to inform marketing strategies for prospecting, boost retention and loyalty, and improve the overall consumer experience. For example, a credit card issuer can analyze spending data to predict an activity or behavior and provide highly personalized messages, offers and content.
AI Opportunity #4) Operational Efficiency
Generative artificial intelligence can automate routine tasks, freeing up time for financial services marketing teams to focus on more strategic initiatives. These tasks include compiling and organizing information as well as generating routine content at scale (legal, market updates).
This can lead to improved efficiency, cost savings and better use of resources.
More on AI in banking:
- The Care and Feeding of ‘Skinny’ Digital Banking Relationships
- See all of our latest coverage of artificial intelligence
How Banks and Credit Unions Must Mitigate the Risks of Generative AI
If you’re a marketer in the banking field, your mind is likely now racing with all the potential compliance, reputational and ethical pitfalls presented by some of these ideas. Generative artificial intelligence can produce a better customer experience.
However, it can also generate output that is misleading or factually wrong. This must be prevented up front.
The technology also presents risks of biased content (age, gender, ethnicity), and is open to manipulation by outside actors for purposes of unethical or criminal activity. But these challenges can be overcome with the right design and management.
Here are five considerations:
1. Data Selection. To avoid biased, non-compliant or other undesirable outputs, marketing teams must design and train generative AIs with diverse and representative data. Attention must be given to ensure that the data and models conform to applicable laws, such as fair-lending laws and regulations. Teams must regularly audit the AI’s outputs and make adjustments to the data, particularly as the models are optimized over time. It may even make sense to engage an independent third party to verify and validate models’ fairness and compliance.
2. Model Specificity. While we are all increasingly familiar with broad-based AIs like ChatGPT, financial services companies will also need narrowly focused generative AIs that perform very select tasks. For example, compared to a general purpose AI, a model built specifically to generate consumer-facing credit or investment education can help ensure that the content being generated is accurate and relevant — and suitable — for the recipient.
3. Rules and Guardrails. Creating business and compliance-based rules, both for internal teams and external consumers, can help mitigate ethical and legal risks. For example, rules can be established that prevent the AI from inadvertently generating misleading or contradictory outputs.
4. Human Monitoring. While generative AI has the capacity to deliver significant efficiencies, it’s important to continuously monitor and audit the outputs. Consider designating a monitoring team — of humans! — to refine, permit or suppress AI-generated outputs. Use that monitoring to test and further train the AI.
5. Ongoing Training. This is going to be an ongoing set of challenges, not a one and done. All stakeholders need to be educated on the capabilities, limitations and risks of the AI. Teams that directly use or who provide data, content and other input into the models should thoroughly understand how the system works and what it is designed to do. Leaders overseeing the AI should be transparent about potential risks, and encourage questions as well as solicit any concerns from stakeholders.
Generative AI, when combined with CRM, can offer financial services companies significant marketing advantages and opportunities. The key will be to balance the technology with the human element, both in the development of the initial AI models, as well as the management of outputs for consumer or business consumption.
At the moment, we are only scratching the surface.
About the author:
Matt Regan is vice president of customer strategy and financial services vertical lead at the customer experience company Merkle.