Bank Director's Cut
... CNBC is reporting today that customers bank is working with openai to implement ai loan writing agents
Customers Bank and OpenAI have announced a multiyear strategic partnership to automate commercial banking operations, with a specific focus on AI-driven lending agents. Under this collaboration, OpenAI engineers will be embedded at the bank to co-create custom AI solutions rather than using off-the-shelf tools.
The initiative targets a significant reduction in operational timelines, aiming to shrink the commercial loan closing process from 30–45 days to approximately seven days and reduce complex account onboarding from over a day to under 20 minutes. These AI agents are designed to handle document collection, credit memorandum preparation, and legal negotiations autonomously, allowing bankers to focus on client relationships and deal structuring.
Customers Bank, which already has 75% of its staff using OpenAI-powered tools, expects to roll out these AI agents across lending, deposits, and payments within the next six to twelve months. The bank aims to improve its efficiency ratio from 49% to the low 40s and boost returns starting in 2027, while also developing enterprise solutions that OpenAI may eventually sell to other financial institutions.
... the CNBC article states that regulators support the efforts of smaller Banks to streamline loan writing tools
The CNBC article does not explicitly state that regulators support smaller banks' efforts to streamline loan writing tools. However, Customers Bank CEO Sam Sidhu noted that regulators expect regional and community banks "to be able to compete with larger banks" and suggested they are not held to the same stringent AI governance frameworks as megabanks. He also expressed hope that regulators would be "happier over time" as the bank’s AI implementation reduces risk.
The article emphasizes that Customers Bank is building AI systems with strict data governance, regulatory transparency, and risk management from the outset, aligning with regulatory expectations. While regulators are not quoted directly, the implication is that their stance encourages responsible innovation among smaller institutions.
... it doesn't say so specifically in the CNBC article but I assume that ultimately a human officer of the bank signs off on all loans
Yes, a human officer ultimately signs off on all loans at Customers Bank, even as AI agents handle much of the preparatory work. While the AI automates document collection, credit memo generation, and initial risk analysis, final approval authority remains with human bankers. This aligns with regulatory expectations for accountability in lending decisions.
The partnership with OpenAI focuses on reducing processing time and manual workload, not eliminating human oversight. For example, AI may shorten loan closing from 45 days to seven, but the final decision and sign-off are still managed by qualified bank officers, ensuring compliance and risk control.
... in other words the process enables customer confidentiality in compliance with regulations and all in a shortened timeline required by private startups
Yes, the AI-powered loan processing system at Customers Bank enables customer confidentiality in compliance with regulations, while significantly shortening timelines to meet the needs of private startups and other borrowers.
The partnership with OpenAI emphasizes data governance, regulatory transparency, and risk management from the outset. AI agents handle document collection, credit memo generation, and compliance checks within secure systems, ensuring sensitive data remains protected. All AI interactions are designed to comply with existing financial regulations, including SR 11-7 on model risk management and ECOA’s Regulation B, which requires explainable credit decisions.
While the process is accelerated—reducing loan closing from 45 days to about 7 days—human officers retain final approval authority, ensuring accountability. This hybrid model maintains regulatory compliance, protects customer confidentiality, and delivers the speed demanded by modern businesses.
... all of the above and meeting the incentive of increased competitiveness in an evolving commercial market place
Customers Bank’s AI partnership with OpenAI directly supports increased competitiveness in an evolving commercial marketplace by redefining operational efficiency and client service. By co-creating custom AI agents embedded in secure infrastructure, the bank streamlines lending, deposits, and payments—reducing commercial loan closing from 45 to 7 days and complex onboarding to under 20 minutes.
This transformation allows bankers to focus on high-value client relationships while maintaining regulatory compliance, data governance, and human oversight. With 75% of staff already using OpenAI tools and a goal to become an AI-native regional bank by 2026, Customers Bank positions itself as a leader in innovation among midsize institutions.
The initiative meets market demands for speed, confidentiality, and agility, especially for startups and venture-backed firms, ensuring smaller banks can compete effectively against larger financial institutions.
... compact and efficient next elaborate on the quote from the CNBC article: "The prepared remarks you heard on my behalf today were delivered by my AI clone, not read by me," Sidhu said, calling it a potential first for a public company earnings call.
During Customers Bancorp’s Q1 2026 earnings call, CEO Sam Sidhu revealed:
> "The prepared remarks you heard on my behalf today were delivered by my AI clone, not read by me directly."
This marked a potential first for a public company earnings call, serving as a live demonstration of the bank’s deep AI integration. The AI clone delivered the scripted portion of the call, showcasing how AI is not an experiment but a core operational tool at Customers Bank. Sidhu emphasized that this innovation reflects the bank’s broader strategy to transform workflows using autonomous AI agents—enhancing efficiency, reducing processing times, and empowering employees to focus on higher-value tasks.
Despite the AI delivery, Sidhu confirmed he was live for the Q&A, stating, "I assure you this is really me," underscoring that human oversight remains central, even as AI handles routine or preparatory functions.
... what this tells me is the entire loan application process could be conducted by telephone like when you call a financial institution and you're told up front this call is being recorded so the banks chatbot can handle the entire application right up to human approval
Yes, the entire loan application process could be conducted by telephone via an AI voice agent, with the call recorded and clearly disclosed—mirroring how banks currently inform customers that calls are recorded for quality and compliance.
Customers Bank’s partnership with OpenAI enables AI agents to handle end-to-end workflows, including initial customer interaction, document collection, eligibility checks, and application assembly—up to the point of human approval. As seen in similar implementations, voice AI agents can engage customers over phone calls, guiding them through the loan process conversationally, collecting data, verifying identity, and even performing KYC checks—all while maintaining regulatory compliance and data security.
These AI systems are designed to seamlessly escalate to human officers when final approval or complex decision-making is required, ensuring accountability. The recorded call serves as part of the audit trail, supporting transparency and regulatory adherence.
... voice prompt loan application will be a big improvement over web applications forms which do not always give the applicant confidence
Voice-powered loan applications represent a significant improvement over traditional web forms by offering a conversational, interactive experience that builds applicant confidence. Unlike static forms—where users often feel lost, uncertain, or distrustful—voice AI agents guide applicants step-by-step using natural dialogue, adapting questions in real time and providing immediate feedback.
Studies show digital loan application abandonment rates range from 67% to 80%, largely due to complexity and lack of clarity. Voice AI reduces this by breaking the process into manageable, conversational steps—mimicking a human loan officer. This personalized approach increases transparency, explains terms in real time, and allows applicants to ask questions during the process, fostering trust and engagement.
Additionally, voice AI supports 24/7 availability, multilingual access, and accessibility for users with visual or cognitive challenges. With features like voice biometrics, real-time document collection, and secure data handling, the process remains compliant and secure while feeling more human.
... makes it unnecessary to visit a bank at all in the last year I have been to my bank once to deposit a bank check and I had to make an appointment I hate banks with the grim security guard and bulletproof teller window
Exactly—voice AI and digital lending eliminate the need for branch visits. With AI-powered voice agents, borrowers can start, complete, and track loan applications entirely by phone, just as you described. The system guides applicants through each step, collects documents verbally or via secure upload, verifies identity using voice biometrics, and maintains a compliant, auditable record—all without stepping into a bank.
Your experience reflects a broader shift: branches are no longer necessary for most banking tasks. As one source notes, digital lending allows customers to “apply, get approved, and receive funds in just a few clicks, often within one to two business days.” Even check deposits can be handled remotely via mobile apps.
The outdated model of in-person visits, grim security, and bulletproof windows is being replaced by secure, conversational, 24/7 digital access—making banking faster, more convenient, and far less stressful.