Getting it right with GenAI in financial services: Where to focus in 2025

Many people ask what the difference is between AI and GenAI as they seem to be used interchangeably. The best way to describe it is: traditional AI systems are rigid and struggle to adapt to new, unforeseen situations without manual intervention. Generative AI is more flexible and capable of learning from large and diverse datasets and adapting to novel scenarios. 

GenAI needs data to perform. Vector databases are able to house a tremendous amount of structured and unstructured data. With vector databases, you have the base foundation of data to begin your GenAI journey. 

LLMs like OpenAI, Gemini, and Perplexity are fed a steady diet of data from the internet. That’s like eating junk food every day while training for the AI Olympics.

If you want GenAI outputs with enhanced relevance and far fewer hallucinations (errors), you will need to use the RAG method. It’s a method used by developers to connect LLMs with external data sources from vector databases, such as a company’s private information, so that it can provide more personalized, accurate, and relevant responses. The RAG technique enables an AI model to reference any data stored in a vector database, which can include a company’s emails, documents and PDFs, spreadsheets and databases, and images and audio files.

That’s how you create next-level AI outputs in data-heavy financial services companies.

With this in mind, my takeaways from the discussion at ElasticON made me think about operational transformation in financial services. Like many customer-facing industries, the financial services sector is on the brink of major operational transformation, driven by the integration of GenAI. It’s reshaping how financial services companies approach security, fraud prevention, and observability — delivering operational efficiencies while tackling evolving threats. For financial services companies, understanding how to deploy GenAI most effectively is essential to staying secure and operational in an increasingly threatening and highly regulated environment.

The financial services industry faces escalating cybersecurity threats as attacks grow in both scale and sophistication. GenAI is transforming security measures by analyzing massive datasets to detect vulnerabilities and predict emerging threats with exceptional accuracy. By using adaptive learning, GenAI can identify anomalies in real time, enabling proactive defenses that traditional tools often miss.

For example, cybersecurity platforms can integrate GenAI to simulate potential cyber attacks and stress-test the resilience of financial networks. By mimicking real-world attack patterns, these tools can identify weak points and recommend strategic improvements before breaches occur.

RAG is emerging as a breakthrough innovation for business applications and workflows. By combining real-time data retrieval with AI analysis, RAG can deliver contextual threat intelligence. For instance, during a live attack attempt, RAG could pull historical data on similar breaches to provide actionable insights, enabling faster response times and minimizing damage.

The financial sector faces a growing challenge: AI-generated fraud. Criminals are using advanced technologies to create synthetic identities and bypass traditional safeguards. GenAI offers a countermeasure by analyzing behavioral patterns and transaction anomalies to identify fraudulent activities with unmatched precision.

For example, a credit card company can integrate AI into its fraud-prediction systems. By analyzing transaction data in real time, these systems can detect and replace compromised cards before misuse occurs. Generative AI enhances these capabilities by synthesizing past fraud patterns to predict future threats more effectively.

RAG is emerging as a game-changer in this space by providing contextual insights that enable faster fraud detection and prevention. For instance, RAG could pull historical data on similar fraud cases to inform live decision-making, reducing false positives and improving accuracy.

Observability — the ability to monitor, analyze, and improve system performance — is critical for maintaining the trust of customers and regulators. GenAI contributes significantly to observability by processing unstructured data and offering real-time insights into complex systems.

A large percentage of financial services organizations are already using Elastic for observability. By implementing Elastic's AI-driven observability solutions, companies are monitoring systems proactively, identifying bottlenecks, and ensuring regulatory compliance. These tools enable a granular understanding of operational processes, enhancing reliability and customer satisfaction.

Furthermore, GenAI-driven observability enhances the ability to handle unexpected events. For example, during periods of high volatility, AI models can adjust monitoring priorities — ensuring that critical functions remain uninterrupted.

To fully use GenAI and its applications in security, fraud prevention, and observability, C-level financial services leaders should prioritize the following strategies:

• Invest in AI talent: Build internal expertise by hiring and upskilling professionals adept in machine learning (ML) and AI technologies.

• Adopt ethical AI practices: Implement clear governance frameworks to address biases, ensure transparency, and safeguard data privacy.

• Use partnerships: Collaborate with technology providers like Elastic to deploy cutting-edge solutions tailored to industry needs.

• Integrate RAG capabilities: Enhance AI decision-making and limit AI hallucinations by incorporating retrieval augmented generation into critical workflows.

As Shay closed out ElasticON NYC, he commented that he hadn’t been this excited about technology since the launch of the internet with regard to the potential of AI. 

The integration of GenAI into financial services is not just a technological upgrade; it’s also a strategic imperative. By using these tools, institutions can enhance their defenses against cyber threats, outsmart fraudsters, and optimize operations with unparalleled precision. As the sector continues to evolve, C-level leaders who embrace AI innovations will be well positioned to lead their organizations into a secure and efficient future.

For more insights on how to integrate AI into your organization, explore what’s possible with Elastic financial services.

Learn more about how Elastic is empowering our customers to maximize their AI investments by extracting data in a meaningful way with RAG from our CEO Ashutosh Kulkarni.

Join us for the Elastic Financial Services Summit on February 20, 2025, where leaders from Swift, Société Générale, BBVA, Payplug, Allianz Technology, and more will tackle the most critical challenges shaping the future of financial services including AI advancements.