Research Finds Two in Three US Banking Executives Believe Agentic AI Will Change How the Industry Operates

While large majority (71%) say they must invest in AI agents to remain competitive, over half (54%) also want stronger regulatory guidance

SoundHound AI today released new research that reveals banking industry attitudes towards agentic AI.

Santa Clara, Calif – October 27, 2025 – SoundHound AI, Inc. (Nasdaq: SOUN), a global leader in voice and conversational AI, today released new research that reveals banking industry attitudes towards agentic AI. The survey of industry professionals, conducted by Arizent, found that 70% predicted the technology will have a “significant” impact or believe the technology will be a “game changer” that will have a profound effect on the future of banking and financial services.

The Current State of Agentic AI in Banking

The research highlights a significant divide in adoption based on institution size, with larger banks leading the charge:

• More than three in five (64%) of larger banks are testing or actively deploying agentic AI.
• Smaller institutions were significantly less likely to be deploying or testing (38%).

Institutions that said they were already deploying agentic AI were almost four times more likely than those in pilot to think of the technology as an industry “game-changer”, with a huge majority (94%) expressing satisfaction with their current agentic AI deployments.

Just 2% of those surveyed said that agentic AI wasn’t at all valuable to their bank.

Key Use Cases

Customer Service was a higher priority for larger banks and senior executives, and more than three quarters (76%) of all respondents said their bank had formed an AI committee to evaluate potential AI use cases.

According to banking professionals, the top use cases for agentic AI are:

• Customer Service / Call Center (59%)
• Employee Efficiency (39%)
• Fraud / Risk Management (37%)

Overall, agentic AI’s core value is seen in its potential to enhance the overall human element of banking, with 73% of respondents believing it will improve the customer experience, and 79% stating it will improve the employee experience.

Adoption Barriers and Incentives

Despite a clear enthusiasm for agentic AI, banks are still navigating a careful path towards adoption. Top concerns include:

• Security and compliance (51%)
• Concerns about autonomy (51%)
• Integration with legacy systems (46%)
• Costs (42%)

Still, many banks are already looking to agentic AI as a part of the solution to the barriers listed above. 42% believe agentic AI could be of benefit to fraud and risk automation, with nearly a third believing that AI agents could help with compliance.

When asked what would make their respective institutions more likely to adopt agentic AI, over half of banking professionals said stronger regulatory guidance (54%) and the ability to integrate seamlessly with existing tools (53%).

SoundHound’s Amelia 7 Agentic AI Platform

Upon being shown a video demo of SoundHound’s Amelia 7 platform, 88% of respondents said that the technology would be valuable to their business, and 98% thought it represented an improvement on their current technology deployment.

“It’s clear that agentic AI will change the banking industry in fundamental ways thanks to its incredible ability to reason, act, and perform complex tasks,” said Michael Anderson, EVP of AI for Enterprise of SoundHound AI. “This independent research illustrates that this evolution is being met with enthusiasm, particularly among larger institutions. However, trust and security have an important role to play in the industry fully embracing agentic AI, and allowing banks to move forward with confidence.”

SoundHound AI works with seven of the top ten global financial institutions and powers hundreds of large enterprise brands – from major retail banks to regional banks and credit unions – with its Amelia 7 platform. The Amelia 7 platform can handle complex, multi-step user queries by orchestrating multiple AI agents with specific functions to answer questions, execute transactions, and resolve problems via text or voice, without the need for human intervention.

Amelia 7’s AI Agents use Agentic+ – a flexible combination of generative AI, multi-agent autonomous orchestration, traditional deterministic models, and human-in-the-loop to complete tasks based on the best possible path to the right response.

Read the full research report here: https://www.soundhound.com/agentic-ai-for-banking-whitepaper-2025/

Methodology

This survey was conducted by Arizent and American Banker on behalf of SoundHound. The research was conducted online from July 22nd to August 18th , 2025 among 201 qualified respondents. To qualify, respondents needed to work at a credit union, retail bank, or commercial bank with a retail division; these banks could be any size. Respondents also had to be in a managerial role or higher at their organization and be at least moderately involved with their organization’s AI initiatives/strategy.

This was a blind data collection effort; SoundHound was not identified as a sponsor of this research.

About SoundHound AI

SoundHound AI (Nasdaq: SOUN), a global leader in voice and conversational AI, delivers solutions that allow businesses to offer superior experiences to their customers. Built on proprietary technology, SoundHound’s voice AI delivers best-in-class speed and accuracy in numerous languages to product creators and service providers across retail, financial services, healthcare, automotive, smart devices, and restaurants. The company’s various groundbreaking AI-driven products include Smart Answering, Smart Ordering, Dynamic Drive-Thru, and the Amelia Platform, which powers AI Agents for enterprise. In addition, SoundHound Chat AI, a powerful voice assistant with integrated Generative AI, and Autonomics, a category-leading operations platform that automates IT processes, have enabled SoundHound to power millions of products and services, and process billions of interactions each year for world class businesses.

Media Contact

Fiona McEvoy

415-610-6590

[email protected]