Banking on agents: These 8 firms are building the future of financial services

Few industries stand to gain more from agentic AI than financial services. It's a sector built on data, trust, speed, and precision, where every decision is subject to regulation and every customer interaction carries real stakes. That makes it a demanding proving ground for AI that does important work: reasoning through dense policy documents, verifying coverage, resolving disputes, and personalizing service for millions of customers.

We’re already seeing a number of companies, from established players to neo banks and fintech startups, implement AI agents.

One is helping companies launch their own credit programs in 25 days, instead of the usual 300. There’s a global card network that’s cut processing times for risk and fraud detection by 90%, while another payments provider built an agent that reduced processing times for complex dispute cases by 70%. A major national bank is reinventing itself around conversational technology. A large insurer can now verify coverage almost instantly and answer policy questions more quickly and accurately.

These applications may vary widely, but they’re built on common foundations: unified, AI-ready data on Google Cloud. From that starting point, most of these companies are using Gemini Enterprise Agent Platform to build and the Gemini Enterprise app to deploy new internal tools and external offerings.

This specialized AI stack is helping organizations across the industry create and operate agents faster, more securely, and with compliance and sovereignty built in. The success we’ve already seen is one of the main reasons we launched Gemini Enterprise for Financial Services earlier this week. And as our ROI of AI research found, agents are where AI's impact are showing up most clearly and quickly.

Below, we look at 8 companies across five continents using AI agents to serve customers, run operations, and equip their teams.

Making money management more personal

For customers, the clearest sign of this shift is AI they can easily interact with, including delivering credit decisions quickly.

Starling launched the UK's first agentic AI financial assistant for personal account holders. The in-app Starling Assistant, built with Google Gemini, responds to natural language and carries out banking tasks on the customer's behalf, from setting up personalized savings goals to organizing bill payments. It also offers tailored support for customers with higher accessibility or socio-economic needs.

“Generative AI has transformative potential for financial services and Starling is showing what’s possible, whether it’s protecting people from scams or helping them better understand their spending.” – Raman Bhatia, CEO, Starling Group

Bullla, a Brazilian fintech with more than 3.6 million cards issued, offers credit, benefits, and salary advances to people historically overlooked by traditional financial institutions. After modernizing its data foundation on BigQuery, Bullla now uses AI for critical processes and regulatory reports that once took hours or days and can now run in minutes, a 90% performance gain. The company also builds and trains the models that assess credit risk on the Gemini Enterprise Agent Platform, allowing Bullla to manage more than R$2.1 billion in credit with greater precision.

Core AI, a São Paulo credit fintech, helps companies launch their own credit programs in 25 days, compared with a market average of around 300. Its automated agents, powered by Gemini models, analyze documents like rental agreements in under 15 seconds, work that once took analysts hours or days. The Gemini Enterprise Agent Platform turns that data into instant credit scores and personalized rates. Operating costs are five times lower than at conventional financial institutions, and the company's risk models keep its default rate about three times below the market average.

Speeding up operations

Behind the scenes, agents are accelerating the machinery of financial services, including highly sensitive areas like fraud detection, disputes, and claims.

Dojo, the UK payments provider processing more than 2 billion transactions a year, adopted Gemini Enterprise across the organization, and employees created more than 680 AI agents within two months. Its Chargeback Agent reduced processing times for complex dispute cases by 70%, sales teams save more than seven hours per person each week, and payment authorization continues to run 60% faster than the industry average.

SIGNAL IDUNA, the German insurer, made Gemini Enterprise its front door to AI and achieved more than 80% monthly active usage in under two months. The platform was rolled out to 11,000 employees, enabling teams across the company to build specialized agents. One example is its Health Agent, which helps service teams automatically verify coverage against a century of complex policy data. The agent has driven a 400% surge in weekly usage and helps deliver customer answers 37% faster.

“We're not just scaling a tool, we're scaling innovation, learning, and the possibilities of what we can achieve.” – Stefan Lemke, CIO, SIGNAL IDUNA

AXA Switzerland, the country's leading insurance provider, built its AI capabilities on a data foundation years in the making. After migrating its data lake to BigQuery in 2021, queries that once consumed an entire weekend now finish in seconds or minutes. That data foundation powers an assistant-first approach to AI. Specialists who spent 25 minutes reading complex medical documents now get key information in seconds; automated triage identifies which claims require expert evaluation; and a fraud detection ruleset uncovered a 120,000 Swiss franc false claim in its first week. With Smart Data projects ranking among the most profitable initiatives in the company's history, AXA is now exploring the next step: agentic AI for specific insurance workflows, along with voice bots and agent assist for its call center and claims teams.

"Cloud transformation has fundamentally changed how we approach IT. We're no longer just implementing business requirements — we're driving innovation and suggesting new possibilities to the business." – Kathrin Braunwarth, Head of DTI (Data, Technology and Innovation), AXA Switzerland

Empowering employees with AI tools

The most ambitious deployments put agent-building capabilities in the hands of the entire workforce.

PayPal set out to build an AI-native engineering platform to accelerate innovation across its trusted global payments infrastructure and power the next generation of commerce. The groundwork started with migrating more than 300 petabytes of data to BigQuery. Today, the company runs its AI workloads on Google Kubernetes Engine and Gemini Enterprise Agent Platform, serving more than 500 production models that handle nearly 1 billion requests a day at 99.999% availability. Model deployment that once took weeks now takes minutes, and PayPal's 10,000 engineers are building agentic commerce capabilities for more than 439 million consumers and merchants.

"Faster innovation, trust at scale, and an entire workforce empowered with AI — that's what building on Google Cloud has made possible for PayPal." – Srinivasan Manoharan, Director of AI/ML Platform, PayPal

Unicaja, one of Spain's largest banks, is partnering with Google Cloud to accelerate its AI strategy toward a Conversational Banking model. This initiative aims to transform customer and employee relationships, making them simpler, closer, and more natural, improving the customer experience, promoting digitalization and inclusion, and increasing operational efficiency. Google will contribute its advanced AI infrastructure capabilities from Google Cloud, Gemini Enterprise Agent Platform, the Agent Development Kit (ADK) framework, responsible AI and governance frameworks, and its ecosystem of AI partners to accelerate AI adoption and transformational use cases.

"Our alliance with Google as a strategic partner allows us to use its tools, environments, and capabilities... to develop generative AI within the organization." – Estrella Botas, Chief Technology and Operations Officer, Unicaja

From pilot to production

Across banking, payments, insurance, and credit, these eight companies have moved AI agents out of experimentation and into daily operations, where regulated, high-stakes work gets done. The results justify their efforts, from faster fraud detection to credit decisions in seconds.

If you're weighing where agents could take on real work in your organization, Gemini Enterprise for Financial Services is a practical place to start, and the ROI of AI report offers a closer look at how organizations are turning AI investments into measurable returns.