Tax season adds stress to banking systems with a surge in transactions, fraud threats, data complexity and necessary risk controls.

This added stress, including to AI systems, can expose gaps in data quality, Mark Blake, financial services industry practice lead at data management solution provider Stibo Systems, tells FinAi News on this episode of “The Buzz” podcast.

Those gaps to watch for in AI systems include:

• Mismatched data;
• Missing data;
• Lineage issues; and
• Auditability issues.

Financial institutions must ensure “they’re meeting and satisfying the regulators, and they need to be confident that their AI models also stand up to that pressure,” Blake says.

Listen as Blake discusses AI readiness at financial institutions this tax season.

Financial institutions have established cloud partners and now those cloud providers are being tapped for agentic AI, Ravi Khokhar, executive vice president and head of cloud at Capgemini, tells FinAi News on this episode of “The Buzz” podcast.

"The role of cloud has now become an enabler for AI across financial services,” Khokar says.

For example, banks are tapping their cloud providers to host agents, creating processes that are “a little bit more intentional for managing risk, compliance, explainability,” he says.

Netherlands-based bank ABN AMRO tapped Microsoft and Capgemini to migrate its existing chatbot to the cloud, according to Capgemini’s ” “World Report Series 2026: Cloud in Financial Services.” The bank migrated its chatbot infrastructure to Microsoft Copilot Studio to create an agent that integrates generative AI features and scalability.

Before moving it to the cloud, the chatbot had high customer drop-off rates, limited capabilities and high operating costs, according to the report. Now, the agent’s language processing accuracy in Dutch has increased by 7% and overall drop-off rates fell.

Listen as Khokhar discusses agentic readiness, multicloud and multiagent ecosystems and the importance of governance, trust and training when deploying agentic AI.

Bank of America invests about $4 billion annually in AI and other technologies to keep up with the needs of its clients.

“AI at the bank … touches everything we do,” Nikki Katz, head of digital at Bank of America, tells FinAi News on this episode of “The Buzz” podcast.

Katz leads a team of designers, data scientists, product managers and others who meet the needs of the 59 million clients accessing the bank through its digital channels — its website, mobile app, and its AI-driven assistant Erica.

“We make over a thousand different enhancements to [the digital experience] every year,” she says, noting, “We’re constantly tuning Erica.”

Erica adoption continues to grow as it is continuously iterated. The latest stats:

• About 50 million clients have used the AI assistant;
• Clients have engaged with Erica for almost 19 million hours; and
• Erica has had over 3 billion interactions.

Listen as Katz discusses how her team approaches emerging technology and the application of AI across digital channels at the bank.

Check fraud remains a threat for financial institutions, but AI can help by detecting anomalies.

Check fraud is “a growing concern for many,” Ati Azemoun, vice president of business development at Longmont, Colo.-based software provider ParaScript, tells FinAi News on this episode of “The Buzz” podcast.

In fact, 63% of organizations experienced check fraud or attempted check fraud in 2025, according to the Association of Financial Professionals. The AFP surveyed 521 treasury practitioner members.

One anomaly AI can detect is signature forgery, Azemoun says.

ParaScript’s forgery detection software, SignatureXpert.AI, can compare signatures from previous checks and flag when a signature is different, ultimately stopping the fraud, he says.

Founded in 1996, ParaScript banking clients include financial institutions, government agencies and companies in the U.S., Europe, Latin America and Australia. In 2025, ParaScript partnered with tech provider BIRGER which will introduce the technology to commercial banks in Mauritius, Africa.

Listen as Azemoun discusses the current state of check fraud and how AI is reducing fraud at financial institutions.

AI, in some capacity, has been used within payments for 30 years. The latest evolution in AI is through the agentic lens — transforming transactions and experiences alike.

“We're actually thinking about really reimagining not one payment, but an entire experience,” Zachary Aron, principal, Deloitte Consulting, tells FinAi News on this episode of “The Buzz” podcast.

For example, how can agents be used for travel, business spending or even date nights, he says, noting that you bought the airfare, but do you need to add a hotel or sightseeing package? Agentic AI can facilitate those payments.

Agentic AI makes payments “invisible and easy,” Aron says.

Listen to “The Buzz” as Aron discusses the possibilities for agentic transactions this year.

Rabobank and Santander Brazil are seeing efficiency gains from generative AI.

The European banks use AI Pega GenAI from workflow automation system provider Pegasystems, Steve Morgan, global banking industry lead at Pega, tells FinAi News on this episode of “The Buzz” podcast.

Financial institutions are implementing AI at scale, but logistics should be the focus before diving headfirst into emerging technology.

Mac Thompson, chief executive of software provider White Clay, tells FinAi News the eight steps he shares with financial institution clients when approaching AI on this episode of “The Buzz.”

Fraud is on the rise, and AI is contributing to both the solution and the problem.

“Fraud is a really large problem and growing exponentially,” Austin Capital Bank Chief Executive Erik Beguintells FinAi News in this episode of “The Buzz” podcast.

Consumers reported losing $12.5 billion to fraud in 2024, according to the FTC.

And that’s probably understated, Beguin says, noting that not all consumers report being defrauded, due to embarrassment, being in denial or other reasons.
Fraudsters are using AI to target individuals, using it to read social profiles and identify the best way to attack a person, he says. This streamlines an extremely laborious task for bad actors.

On the other hand, FIs are using AI to boost account security, Beguin says. For example, consumers and financial institutions are likely to start moving away from usernames and passwords and even traditional multifactor authentication methods. Instead, the market should move toward biometrics and AI-driven device and image authentication.