Key Facts

  • Company: NatWest Group
  • Company Size: Serves over 19 million customers
  • Location: United Kingdom (Edinburgh headquarters)
  • AI Tool Used: OpenAI generative AI LLMs in Cora+ and Ask Archie
  • Outcome Achieved: 150% increase in Cora customer satisfaction scores (2024)

Want to achieve similar results with AI?

Let us help you identify and implement the right AI solutions for your business.

The Challenge

NatWest Group, a leading UK bank serving over 19 million customers, grappled with escalating demands for digital customer service. Traditional systems like the original Cora chatbot handled routine queries effectively but struggled with complex, nuanced interactions, often escalating 80-90% of cases to human agents. This led to delays, higher operational costs, and risks to customer satisfaction amid rising expectations for instant, personalized support [2].

Simultaneously, the surge in financial fraud posed a critical threat, requiring seamless fraud reporting and detection within chat interfaces without compromising security or user trust. Regulatory compliance, data privacy under UK GDPR, and ethical AI deployment added layers of complexity, as the bank aimed to scale support while minimizing errors in high-stakes banking scenarios [5][6].

Balancing innovation with reliability was paramount; poor AI performance could erode trust in a sector where customer satisfaction directly impacts retention and revenue [7].

The Solution

Cora+, launched in June 2024, marked NatWest's first major upgrade using generative AI to enable proactive, intuitive responses for complex queries, reducing escalations and enhancing self-service [2]. This built on Cora's established platform, which already managed millions of interactions monthly.

In a pioneering move, NatWest partnered with OpenAI in March 2025—becoming the first UK-headquartered bank to do so—integrating LLMs into both customer-facing Cora and internal tool Ask Archie. This allowed natural language processing for fraud reports, personalized advice, and process simplification while embedding safeguards for compliance and bias mitigation [1][3].

The approach emphasized ethical AI, with rigorous testing, human oversight, and continuous monitoring to ensure safe, accurate interactions in fraud detection and service delivery [4][7].

Quantitative Results

  • **150%** increase in Cora customer satisfaction scores (2024)
  • **Proactive resolution** of complex queries without human intervention
  • First UK bank OpenAI partnership, accelerating AI adoption
  • Enhanced fraud detection via real-time chat analysis
  • Millions of monthly interactions handled autonomously
  • **Significant reduction** in agent escalation rates

Ready to transform your business with AI?

Book a free consultation to explore how AI can solve your specific challenges.

Implementation Details

Background and Evolution of Cora

NatWest's digital assistant Cora was initially launched years ago to streamline customer interactions, but by 2024, it needed evolution to meet growing demands. The bank identified limitations in managing complex queries like personalized financial advice or intricate fraud reports, prompting the development of Cora+.[2] This upgrade introduced generative AI capabilities, enabling more natural conversations and proactive suggestions.

Complementing Cora, the internal assistant Ask Archie was enhanced to support employees, reducing internal query resolution times and freeing resources for customer-facing improvements.

Strategic Partnership with OpenAI

On March 20, 2025, NatWest announced its collaboration with OpenAI, a milestone making it the first UK-headquartered bank to partner directly with the AI leader. The initiative focuses on leveraging OpenAI's LLMs (large language models) to power digital assistants, enhance customer support, and drive bank-wide simplification.[1][3] Key objectives include combating financial fraud through intelligent chat analysis and delivering hyper-personalized service.

The partnership extends beyond chatbots, integrating AI into broader processes like fraud prevention and operational efficiency, aligning with NatWest's strategic AI research efforts.[6]

Technical Integration and Architecture

Implementation involved fine-tuning OpenAI models on NatWest's vast dataset of anonymized interactions, ensuring domain-specific accuracy for banking terminology, regulations, and customer behaviors. Retrieval-Augmented Generation (RAG) techniques were employed to ground responses in verified data, minimizing hallucinations. Security layers included prompt engineering for compliance, real-time monitoring, and human-in-the-loop escalation for sensitive topics like fraud.[4]

For fraud detection, the LLM analyzes conversational cues (e.g., urgency, inconsistencies) to flag risks and guide users through reporting seamlessly. Internal Ask Archie uses similar tech to assist staff in verifying fraud claims faster. The system processes millions of interactions monthly, with AI handling initial triage.[5]

Implementation Timeline and Approach

The rollout began with Cora+ pilot in early 2024, gathering feedback before full deployment in June. Post-partnership, OpenAI integration rolled out in phases: Q2 2025 for testing, Q3 for customer beta, and full production by Q4. NatWest adopted an agile methodology with cross-functional teams (data scientists, compliance experts, UX designers), conducting A/B testing to measure uplift.[1]

Challenges like model bias and latency were overcome via ethical AI frameworks, continuous retraining, and partnerships for specialized banking LLMs. Employee training programs ensured smooth adoption of Ask Archie.

Safety, Ethics, and Ongoing Optimization

NatWest prioritizes responsible AI, with guidelines from their Chief Data & Analytics Officer emphasizing transparency and fairness. Regular audits, customer feedback loops, and adversarial testing safeguard against misuse in fraud scenarios.[7] Metrics tracking includes CSAT, resolution rates, and fraud detection accuracy, with iterative updates based on real-time data.[6] By late 2025, the system continues to evolve, positioning NatWest as an AI leader in UK banking.

Interested in AI for your industry?

Discover how we can help you implement similar solutions.

Results

Following the Cora+ launch and OpenAI integration, NatWest achieved a dramatic 150% increase in customer satisfaction scores for Cora in 2024, as more queries were resolved proactively without human intervention.[2] The generative AI enabled handling of complex issues like multi-step financial advice and fraud reports, boosting self-service adoption and reducing average resolution times significantly.[4] Fraud detection saw marked improvements, with chat-based reporting streamlined for faster triage and pattern recognition, aligning with UK banking trends where AI cuts fraud losses.[5] Internal metrics show Ask Archie accelerating employee productivity, contributing to overall cost savings estimated in millions through fewer escalations. By December 2025, the partnership has solidified NatWest's position as an innovator, with millions of enhanced interactions monthly and sustained high CSAT. Ongoing research ensures ethical scaling, with future expansions into personalized banking products.[1][3][7] This case exemplifies how LLM integration transforms legacy banking services into intuitive, secure experiences.

Contact Us!

0/10 min.

Contact Directly

Your Contact

Philipp M. W. Hoffmann

Founder & Partner

Address

Reruption GmbH

Falkertstraße 2

70176 Stuttgart

Social Media