Enterprise AI has spent the past two years proving its ability to summarize documents, answer questions, and automate repetitive work. Banking spent the same two years asking a different question: Can AI be trusted to make decisions that carry financial, regulatory and legal consequences?
That threshold, as well Mike Taro Wimmerco-founder and CEO of Touchshe told PYMNTS CEO Karen Webster that she was starting to take action.
“The models were not good enough to be ready to make critical decisions related to financial services,” he said. “2026 is the year AI will come to financial services.”
Following a $110 million funding round led by Goldman Sachs Alternatives, Taktile is betting that the market for independent financial decision-making is arriving faster than many executives realize. The company’s platform deploys AI agents specifically designed for regulated financial institutions, helping banks automate some of their most complex operational workflows while maintaining regulatory oversight.
Wehmeyer believes the financial services industry is moving toward an “agent-first” future where conversational AI becomes the primary interface for opening accounts, applying for loans, and interacting with financial institutions.
The bet is less about replacing humans and more about redefining how organizations interact with AI.
The real inflection point for AI is better risk decisions, not better models
Despite the hype and headlines surrounding AI, many enterprise executives remain reluctant to give AI agents access to core enterprise systems, let alone allow them to make high-stakes decisions. This is especially true for highly regulated industries such as banking.
Webster opened the discussion by asking the central question facing companies building enterprise AI today: “Are you betting on a market that exists today or a market that you see evolving and accelerating?”
For Weimayer, the answer was unequivocal. Rather than simply assisting employees, AI can now autonomously complete sophisticated workflows across complex financial areas such as commercial lending, insurance claims management, and business underwriting that previously required experienced analysts.
“We are betting on the market that this will be possible,” he said, acknowledging that while thousands of banks remain cautious, a growing group has already moved beyond experimentation into production.
Using AI solutions, a small business loan that previously required weeks of manual underwriting can be approved in a matter of minutes, while insurance claims that historically took months can be evaluated in a matter of hours using drone imagery and AI-powered damage assessments.
“I think a lot of people are now confusing AI transformation with cost savings,” Wehmeyer said, noting that the biggest competitive advantage now comes from AI’s ability to significantly compress decision times.
“If I’m a small business owner and I’m asking for a loan, and I don’t get a response within 14 days…but within five minutes, how great would that be?” He added.
For companies operating under cash flow pressure, speed of decision-making often becomes as valuable as capital itself. As a result, competitive differentiation shifts from those offering financial products to those offering financial certainty more quickly. This distinction transforms AI from production software into operational infrastructure.
Leadership matters more than organization size when it comes to AI readiness
One surprising lesson emerging from Taktile’s customer base is that AI readiness has nothing to do with organizational size. While some of the industry’s largest banks remain hesitant to embrace autonomous AI, some community banks and credit unions have become aggressive adopters after completing broader cloud modernization efforts.
This has a leveling effect. AI models capable of underwriting a business loan or clearing a KYB check were, until recently, the domain of the largest organizations with the R&D budgets and technical teams to build them. Taktile puts the same decision-making power in the hands of small banks and credit unions. A community institution can now compete for the same customers, deposits, and small business relationships as a national bank, without having to set up a data science organization to do so.
For a smaller financial institution, this is the difference between watching larger competitors compress a two-week loan decision into five minutes and being able to do it themselves. Speed of decision making becomes a product that a $2 billion credit union can deliver just as easily as a $2 trillion bank.
As Webster notes, organizations must adapt not only business models but also employee behavior.
“The technology is amazing,” she said. “It’s change management, getting people comfortable with this powerful technology.”
This reflects an often overlooked fact regarding AI adoption in organizations. Deploying autonomous systems requires more than just investment in technology. It requires organizational readiness to redesign decision-making processes, redefine employee responsibilities and accept new operating models.
“It’s not just about what’s possible,” Weimayer said. “The question is how quickly can we get there?”
To reduce barriers to adoption, Taktile has created its own research organization, Taktile Labs, dedicated to measuring the performance of AI against human experts across financial use cases. Taktile Labs serves as the evidence engine for the company. It publishes ongoing data on how model performance is evolving across underwriting, KYB, fraud and claims, giving organizations a continuous read on where AI has bridged the gap with human analysts and where it has not yet. For the risk officer who is asked to hand the decision over to the machine, this shift from vendor assertion to measured and published criteria is often what moves the conversation forward. The company also encourages customers to run AI in “shadow mode,” allowing organizations to compare AI recommendations with existing human workflows before granting systems operational authority.
This gradual shift reflects a broader shift happening across enterprise AI. Organizations no longer wonder whether AI works or not. Instead, they question whether the company’s performance is consistent enough to withstand audits, regulatory scrutiny and executive accountability.
The first agent bank is already taking shape
Looking to the future, Webster posed a provocative question: Could banks eventually become infrastructure behind AI agents rather than core customer interfaces?
“I expect banks to be very agent-first and API-first,” Weimayer said.
Humans are likely to remain involved in making the largest and most important decisions, especially when regulators require oversight. But routine financial interactions may increasingly occur between intelligent software systems rather than between people.
“The technology is going to be there,” Wehmeyer said. “The question is… when will the regulator be ready for this?”
This observation reflects the broader reality facing AI in enterprises. The technology race is increasingly giving way to the trust race. The organizations defining the next decade of AI may not simply be building the smartest models. They will be the ones who convince organizations that these models deserve a seat at the decision-making table.
a witness The full PYMNTS “Monday Conversation” interview. With Taktile CEO Mike Taro Wehmeyer to hear more about:
- Why the biggest barrier to enterprise AI is no longer capability but trust.Recent advances in AI models have made autonomous underwriting, KYB decisions and insurance claims technically possible, but widespread adoption now depends on proving that those systems can command the trust of banks, regulators and executives, Wehmeyer says.
- How competitive pressure – not just cost savings – is accelerating the adoption of AI in banking.The discussion explores why faster loan approval processes, real-time underwriting, and dramatically shorter claims processing have become strategic differentiators that improve the customer experience while helping financial institutions compete for businesses and consumers.
- Why financial AI winners will combine specialized software with organizational transformation. Wehmeyer argues that success requires more than just strong models, stressing that industry-specific expertise, regulatory-ready infrastructure, rigorous standards and practical change management are becoming the real moats as banks move AI agents into production.
