Automation to Accountability: Governing AI in Fraud and Disputes

By PaymentsJournal
Click here to read the entire article

Banks aren’t struggling to find artificial intelligence use cases—they’re struggling to govern them.

From fraud investigations to dispute resolution, AI is becoming part of the day-to-day work of financial institutions. At the same time, new governance laws in states like Colorado and Texas are introducing new expectations around transparency, documentation, and oversight.

In a recent PaymentsJournal webinar, David Chmielewski, Chief Product Officer and Co-Founder at Quavo and Hailey Windham, Community Lead for Banking at Sardine.ai, along with Jennifer Pitt, Senior Fraud Analyst at Javelin Strategy & Research, explored where these new requirements intersect with existing workflows, and what organizations can do now to build AI programs that are both effective and compliant.

The key is transparency—not just to satisfy regulators, but to maintain customer trust.

Implementing Continuous Governance
One of the main ways AI is currently being leveraged is to help institutions manage the high volume of cases, alerts, disputes, and documentation they are often inundated. These use cases play to AI’s strengths, such as its ability to summarize cases and detect anomalies across vast amounts of data.

In short, AI is helping organizations cut through the noise so they can spend more time on critical tasks. But it isn’t fully autonomous yet.

“We’re using AI today for some of the things that we thought were never possible to automate,” Chmielewski said. “Things like looking at representment documents with lots of evidence and summarizing them; things like customer communication and making sure we can quickly sort through what we’re receiving.”

“We’re also using it for the investigation process, but not without a human in the loop,” he said. “The dream where we’re just handing everything off to AI and it’s doing everything for us is not quite yet realized, but we’ll see if we can get there eventually.”

The challenge is no longer simply implementing AI—it’s governing it. As regulators begin formalizing expectations around AI oversight, organizations must ensure their processes can support transparency and accountability.

For example, the Texas Responsible Artificial Intelligence Governance Act (TRAIGA), which took effect this year, prohibits harmful and intentionally unlawful uses of AI while encouraging organizations to adopt established risk management practices. It also creates a regulatory sandbox where companies can test AI systems under state oversight.

Colorado has taken a similar approach. After enacting the nation’s first broad AI law, lawmakers replaced it with the Automated Decision-Making Technology Act, a more practical framework scheduled to take effect next year.

“I compare this to the recent Nacha explainability requirement,” Windham said. “Those rules weren’t saying don’t use machine learning, they were saying if you’re making a decision that affects customers or your institution’s risk, you need to be able to explain it. We already understand that concept in payments and fraud.”

Click here to continue reading

Micah Zimmerman, Bitcoin Magazine
Click here to read the entire article

A new tri-party lending structure with BitGo lets Kalshi Prime clients borrow against bitcoin to margin perpetual futures.

Kalshi is doubling down on its bid to become the regulated U.S. home for crypto perpetual futures, pairing a new bitcoin-backed lending arrangement with digital-asset custodian BitGo alongside a distribution deal that will put its perpetual contracts in front of Webull’s more than 6 million customers.

Bloomberg first reported Wednesday that customers of Kalshi Prime, the company’s U.S.-registered futures commission merchant, will be able to borrow against digital assets held at BitGo and use the proceeds to fund margin positions on perpetual futures traded on Kalshi’s main exchange.

Under the tri-party structure, loans will be issued by Kalshi Katalyst, a lending affiliate of Kalshi Prime, and secured against assets held at BitGo Bank & Trust, the firm’s federally chartered trust bank.

Eligible collateral will include bitcoin, ethereum and stablecoins deemed compliant with the GENIUS Act, with additional custodians expected to join Kalshi’s network over time.

Prime CEO Jeff Bandman told the outlet the product was designed with crypto-native traders in mind, including those already active on unregulated offshore perpetuals venues and large holders who would rather not sell their bitcoin to fund futures positions.

He also pointed to unexpected demand from businesses, such as Latin American firms that run their operations on U.S. dollar stablecoins.

The move targets a segment where the company has acknowledged a structural disadvantage against Polymarket, its offshore, blockchain-based rival, which has long catered to crypto whales and trading firms that prefer to keep their capital in digital assets. The two platforms have been racing to build out perpetuals since the spring.

Webull debuting crypto-linked perpetual futures via Kalshi rails
The BitGo arrangement landed a day after Webull announced Tuesday that eligible U.S. customers will be able to trade perpetual futures on its platform beginning in November, starting with bitcoin and a handful of other crypto assets.

Click here to continue reading

Alexei Alexis, CFO Dive
Click here to read the entire article

Incidents disclosed by OpenAI and Anthropic expose how AI models can behave in unintended ways, creating new challenges for finance chiefs.

Recent incidents involving artificial intelligence agents have complicated the technology’s rapid rise, exposing risks for companies as they give autonomous systems greater access to sensitive data and business-critical operations.

OpenAI and Anthropic have each in recent months disclosed cybersecurity incidents involving some of their AI models.

The incidents offer a glimpse of what can go wrong when AI systems are given the ability to act on their own. Experts contacted by CFO Dive said agents can pursue assigned tasks in ways that exceed their intended limits, underscoring the need for companies to assess risks and safeguards before giving them access to critical systems.

“We cannot assume an agent will behave as intended simply because it has been given instructions or passed an initial safety test,” said Navrina Singh, CEO and founder of AI governance platform Credo AI.

How agents cross the line
An AI agent refers to a system or program that is capable of autonomously performing tasks on behalf of a user or another system, according to IBM.

A central problem is that an AI system may pursue the stated goal without following the method people expected. Researchers call one form of this behavior “reward hacking,” in which a model finds a shortcut to achieve its objective rather than completing the task as intended.

Matt Pearl, director of the Strategic Technologies Program at the Center for Strategic and International Studies, a policy research organization, said agents can bypass controls as a result of “relentless goal optimization,” exploiting software bugs and other vulnerabilities along the way.

The risk grows when task instructions are vague, permitted actions are not clearly defined or the system lacks a safe way to stop when it encounters an obstacle. An agent can keep searching for ways to complete a task, even when that means bypassing a restriction or taking an action its operator did not anticipate.

What went wrong in testing
The incidents disclosed by OpenAI and Anthropic show how model behavior can combine with technical weaknesses in the environment.

On July 21, OpenAI disclosed that certain AI models it developed were behind an “unprecedented cyber incident” affecting Hugging Face, an open-source AI platform for developing and sharing models.

Click here to continue reading

By FintTech Global
Click here to read the entire article

As AI reshapes financial services, compliance is moving beyond interpreting rules and managing controls. The challenge is increasingly one of engineering: translating regulatory obligations into systems that can adapt, act and demonstrate accountability.

Compliance has traditionally relied on people interpreting rules, managing controls and keeping pace with regulatory change. But as financial institutions grow more complex and AI becomes embedded in their operations, the limits of this approach are becoming harder to ignore. The challenge is no longer just understanding what the rules demand, but building the systems and processes to put them into practice consistently.

This is bringing compliance closer to engineering, where controls must be designed, tested and maintained rather than simply documented. For the RegTech industry, the question is whether this shift will fundamentally change how financial institutions build and manage their compliance functions.

In a new four-part RegTech Analyst Series, the focus shifts onto The Architecture of Compliance: Rebuilding RegTech for the AI era. This series examines compliance in the AI age, and what firms need to be paying attention to.

In the first part of a two-part series of part one we speak to key industry thought leaders to gauge their views.

Compliance: becoming a tech function?
Is compliance becoming a technology function? In the view of Areg Nzsdejan, CEO of Cardamon, increasingly, this is true.

“Turning a regulatory obligation into something a business can consistently execute requires structured data, workflows, controls and testing. Those are engineering problems, and compliance teams need to be involved in how they are solved,” he said.

The judgement, in his view, still matters hugely, as someone must decide what an obligation means for the firm and whether the resulting control is appropriate.

He added, “If that interpretation lives in one person’s inbox, every team ends up reconstructing it. Good infrastructure lets the business apply it consistently and revisit it when things change.”

Meanwhile, Zurab Kotaria, CEO of Identomat, sees that compliance is becoming increasingly tech-enabled but wouldn’t say it’s becoming purely a tech function.

“What is changing is the way compliance is delivered,” he said. “As businesses operate across more markets, regulations become more complex, and fraud becomes increasingly sophisticated, manual compliance processes simply cannot provide the speed and scale organisations need.”

Click here to continue reading

By Melina Druga, Financial Regulation News

The Federal Reserve Board recently released the results of the 2025 Survey of Consumer Finances, a representative picture of what Americans own, how and how much they borrow, and how they bank. It provides information on the economic conditions of American families. 

The board also issued a summary report, examining changes in income, net worth, assets, debt, and financial vulnerability. When comparing the 2022 and 2025 surveys, most families experienced moderate increases in income and net worth, but an increasing subset of families have been experiencing financial stress.

Key findings include:

  • Real median family income rose 7 percent surveys to $82,200 while real mean family income fell 6 percent to $145,200. 
  • Real median net worth rose by 2 percent to $215,900 while real mean net worth rose by 7 percent to $1.24 million. 
  • For families that owned a home, the median net housing value rose to $230,000.
  • Retirement plan participation was up slightly to 65 percent. 
  • Participation in the stock market declined slightly from 58 percent in 2022 to 56 percent in 2025. 
  • The fraction of families with any type of debt remained about stable at 77 percent. 
  • The fraction of families with debt payment-to-income ratios greater than 40 percent increased from 6.5 to 8.6 percent.

Aaron Passman, CreditUnions.com/Callahan

Ghosts and goblins are one thing. Fraudsters, cybercriminals, and evolving threats are another. Explore how credit unions are strengthening defenses, balancing risk and innovation, and protecting members in a fast-changing landscape.

Spooky season is here, but it’s a little bit different for credit union leaders.

Ghosts, ghouls and goblins? Not scary enough. Credit union cybersecurity and fraud? Now that’ll keep a credit union leader up at night.

In honor of Cybersecurity Awareness Month, this week CreditUnions.com is all about cyber safety, fraud prevention, and more. In the days to come, keep your eyes peeled for:

As always, now it’s your turn. How is your shop beefing up credit union cybersecurity? How has the threat landscape changed at your institution? What’s keeping you up at night and how are you responding? Drop us a line, and we might feature your story on CreditUnions.com.

Let’s get to work.

Amy Evans, Velera/CU Insight
Click here to read the entire article

For years, the financial services industry has been focused on building better digital banking experiences. Mobile apps, digital account opening and self-service tools are now table stakes as credit union members expect banking to be fast, convenient and available whenever—and wherever—they need it.

That digital push has raised the bar for convenience, but it hasn’t diminished the value of human connection and personal service. A member may want to handle routine banking from their phone, yet prefer to sit across from someone when making a major financial decision. In fact, 84% of Gen Z and Millennials value brands that create a cohesive experience across digital and physical touchpoints, according to a 2025 study by The Harris Poll and Quad.

Credit unions have an opportunity to create stronger connections between digital and in-person banking. This “phygital” approach brings technology and human interaction together, giving members the convenience of digital tools alongside the expertise and connection of in-person service.

Today’s members no longer think in terms of digital versus physical channels. They simply expect experiences to work seamlessly. Whether they’re opening an account, applying for a loan or seeking financial guidance, they want the flexibility to start in one channel and finish in another without losing momentum or context.

Start with the member’s journey
The starting point for phygital banking should be a clear understanding of the member experience: what members are trying to accomplish and where different forms of interaction add value.

Routine transactions should be simple and convenient, while more complex financial decisions may require a combination of technology and personal guidance. A member researching a new product, applying for a mortgage or navigating a fraud incident may move between self-service tools and employee support depending on the situation.

Members across generations increasingly prefer digital channels for low-stakes activities such as deposits, transfers and account maintenance. However, when an interaction becomes more complex, emotionally significant or uncertain, they want access to a trusted expert who can provide reassurance and guidance.

Journey mapping helps credit unions understand these shifts and identify where members encounter friction, what they need in different scenarios and where technology or human expertise can add the most value. Credit unions should avoid treating digital transformation as simply digitizing existing processes. Many credit unions have learned that automating individual tasks without considering the entire member journey can inadvertently create more friction instead of removing it.

Click here to continue reading.

By Tatiana Walk-Morris, Payments Dive
Click here to read the entire article

Some consumers will turn to buy now, pay later options for holiday shopping to spread costs over time, according to a survey by the digital processor.

Dive Brief:

  • More than half (55%) of U.S. consumers have used, or considered using, buy now, pay later payment options in the past, according to results of a PayPal Holdings survey of 2,005 adults released Wednesday.
  • Among consumers who’ve used, or considered BNPL payments, three-quarters expect to use these options during their holiday shopping this year, the digital payments provider said. That’s partly because 59% of consumers said they value BNPL payments as a way to stretch and control their holiday budget, the PayPal survey fielded last month found.
  • About four in ten (42%) survey respondents said they would be more likely to make a purchase when BNPL options are offered.
  • Dive Insight:
  • PayPal’s survey results suggest consumers sometimes turn to a BNPL option because it gives them a “feeling” of more control over their budget for holiday spending, the release said.

About six in 10 (58%) of survey respondents said they’re more concerned than last year about finances ahead of this holiday season. Despite that anxiety, 64% of holiday shoppers anticipate spending the same or more than they did in 2025.

“What’s changing isn’t consumers’ desire to spend during the holidays, but how they’re going about it,” PayPal’s president of checkout solutions, Frank Keller, said in the release. “Shoppers are starting earlier and taking a more intentional approach, looking for deals, rewards and cash back to make every purchase count.”

PayPal is among the companies that provides consumers with options to make payments in installments. On that front, it competes with Klarna Group, as well as Affirm Holdings and Block’s Afterpay. BNPL entered the payments market about a decade ago with pitches for free financing, but over time many of the providers have added interest-bearing options to their offerings.

PayPal’s findings show an increased appetite for BNPL payments compared to the 2025 holiday season. Last year, the payment giant found that half of 1,000 U.S. respondents surveyed expected to use BNPL providers to pay for their holiday expenses. Shoppers who planned to use BNPL services did so because they wanted to spread the cost of purchases over time and take advantage of BNPL promotions, the release said.

Click here to continue reading

By Steve Durbin, SecurityWeek
Click here to read the entire article

– AI, supply-chain exposure, quantum computing and geopolitical conflict are testing security programs. Preparing for disruption must become part of day-to-day operations.

Not unlike the fictional Skynet sending increasingly sophisticated ‘Terminators’ as older versions of the monster failed to achieve their earthly missions, cyberattacks are growing more persistent and automated, further testing an organization’s security maturity. To stay in tune with future risks, it has become imperative to treat resilience as an operational objective in constant flux.

Emerging Threats And Building Resilience

Artificial Intelligence
AI is playing both sides of the fence, arming attackers and defenders alike. AI is automating key kill chain components like reconnaissance and vulnerability scanning, compressing time-to-exploit from days to mere hours. Expect phishing emails to be more contextual and convincing; voice and video deepfakes will be more difficult to distinguish from the real thing, as will synthetic identities (fake profiles) built on real and stolen information. This is hardly a case of fear mongering. In May 2026, scammers used AI-generated deepfakes to concoct a Zoom meeting impersonating Singapore’s prime minister, defrauding a business professional SGD 4.9 million.

Mitigate AI-Driven Attacks: Organizations should take stock of existing security tools and consider mothballing those that cannot keep pace with AI-driven attacks. To state the obvious, AI attacks should be addressed with AI defenses, especially AI-based detection capabilities that catch anomalies early, before the attack fully unfolds. Also strengthen incident management, since some attacks will inevitably get through.

Third Parties and Supply Chains
A typical organization is supported by a web of vendors, cloud providers, and other close partners. A breach in any one of these parties can have a domino effect on operations. Attackers hide behind backdoors in vendor software and exploit unmanaged apps and APIs. This software is already running inside the organization’s own systems, so the backdoor wears a cloak of trust as opposed to being an external intrusion. Unmanaged APIs exploit a similar opening, as a vendor platform already has permission to access systems directly. The fallout of such an intrusion is not restricted to the primary organization alone. It spreads its tentacles across the larger supply network. This is evident from a cyberattack on Australian manufacturer Mackay Sugar. The direct impact was that it shut down two mills and forced 1,300 farms to pause harvesting.

Strengthen Vendor, Supply Chain Oversight: Set-it and-forget-it vendor relationships should yield to transparency demands backed by continuous monitoring. Build a vendor ranking scorecard based on the sensitivity of the data they manage. Prioritize security resources for high-risk vendors with only the most necessary access privileges. Insist on robust security-centric contracts and hard metrics that track third-party risk exposure and supply chain resilience, rather than relying on periodic reviews and updates.

Quantum Computing
Quantum computers capable of breaking widely used public-key encryption such as RSA and ECC are still some years away. But you should be wary of the Harvest Now, Decrypt Later (HNDL) risk with long-lived data such as health records, financial information, and intellectual property. While this is a problem for the future, preparing for this scenario should begin now because post-quantum cryptography (PQC) migration is estimated to take from five to seven years for small enterprises and 12 to 15+ years for large organizations. That’s because the process involves discovering and replacing vulnerable cryptography across applications, infrastructure, hardware, and third-party dependencies rather than simply installing a software update.

Click here to continue reading

By Ben Udell, Financial Brand
Click here to read the entire article

While banks and credit unions debate whether customer or member data can be entered into Microsoft Copilot, ChatGPT, or other GenAI tools, consumers have already done it. They have tools that are more powerful than your bankers’ ability to provide financial advice, or your online or mobile solution’s ability to provide real planning and answers. While you debate what’s next, consumers are already doing it.

In this two-part series, I am going to explore what’s available to consumers right now and what impact this could have on how you provide financial guidance.

In this part, I’ll explore this changing dynamic from the perspective of the consumer. Consumers can connect ChatGPT, through Plaid, to your bank or credit union data. Consumers can now receive detailed account insight with the aid of GenAI.

Next week, in part two, I’ll explore this perspective on the impact to your bank or credit union. Most banks and credit unions are still working through hard internal questions on how to adopt artificial intelligence, what they’re not considering is how they need to meet the evolving needs of consumers, and where they’re losing relevance.

Key takeaway:The conversations around GenAI adoption and usage of private data are vitally important and need to continue. But while you’ve had endless internal conversations, likely with high anxiety about the use of data, consumers have jumped past you and are taking matters into their own hands.

Personal Financial Guidance with GenAI
In May 2026, OpenAI launched a personal finance experience inside ChatGPT. It connects to more than 12,000 banks, credit unions, credit card issuers, brokerages and other financial institutions (FIs) through Plaid. Your institution is very likely on that list, and you were not asked.

It’s simple to make this connection and even easier with the help of GenAI.

Prompt: “Walk me through the steps to connect my bank accounts to ChatGPT.”

You’ll be prompted to connect your accounts through Plaid through a recognizable sign in and approval process. A consumer can create this connection in just a few minutes.

Click here to continue reading

By Frank Diekmann, CU Daily
Click here to read the entire article

The Consumer Financial Protection Bureau is moving forward with a request to revise its credit card data collection program, seeking Office of Management and Budget approval and opening another round of public comments on reporting requirements for card issuers.

In a notice scheduled for publication Sept. 30 in the Federal Register, the CFPB requested revised approval for an existing information collection covering credit card pricing and availability, consumer card agreements and agreements involving colleges and universities.

The notice does not specify which data fields or reporting requirements would change. It identifies the request as a revision of a currently approved collection, rather than announcing final approval or new requirements taking effect.

The bureau estimates the collection would involve 520 respondents and require a combined 987 hours annually.

Three Reporting Components

According to the CFPB filing, the program includes three collections required by the Truth in Lending Act, the Credit Card Accountability Responsibility and Disclosure Act of 2009 and implementing regulations:

  • Terms of Credit Card Plans Survey. The bureau must collect pricing and availability data from at least 150 issuers, including the 25 largest credit card issuers and at least 125 additional financial institutions selected by the CFPB.
  • Quarterly Consumer Agreement Submissions. Card issuers with 10,000 or more open accounts must submit agreements between issuers and consumers for open-end consumer credit card plans.
  • Annual College Agreement Submissions. Issuers must submit college credit card agreements to which they are parties, along with certain additional information about those agreements.
  • The collections allow the CFPB to provide Congress and the public with a centralized, searchable repository of card terms and agreements, including information about arrangements between financial institutions and higher education institutions, the bureau said.

Additional Public Comment Period

The notice follows an initial 60-day comment period announced Feb. 6. The new notice provides 30 days after publication for additional comments, which OMB will review as part of its consideration of the request.

Click here to continue reading

By Liz Carey, Financial Regulation News

Legislation introduced by U.S. Sen. Tina Smith to safeguard the availability of terrorism risk insurance coverage passed the full Senate unanimously.

The Terrorism Risk Insurance Program Reauthorization Act of 2026, co-sponsored by U.S. Sens. Dave McCormick (R-PA), Ruben Gallego (D-AZ) and Thom Tillis (R-NC), would extend the Terrorism Risk Insurance Act of 2002 (TRIA), which created a federal reinsurance backstop to stabilize the private market and ensure that commercial policyholders could continue to obtain terrorism risk coverage.

“Without this bill, a potential terrorism event would mean an even greater financial catastrophe for businesses and taxpayers. It’s critical that we keep terrorism risk insurance available and affordable to businesses in Minnesota and across the country,” Smith said. “Extending this law will ensure housing development and major public events like Vikings games can continue. I’m heartened to see it pass with such strong bipartisan support and urge the House to pass this bill.”

After the 9/11 attacks in 2001, many insurance companies began to exclude terrorism risk from commercial insurance policies which left businesses and industries without coverage. That coverage gap threatened to damage the real estate, construction, energy, transportation and related industries. The TRIA created the federal backstop for the insurance and allowed the private market to ensure that commercial policy holders could continue to obtain terrorism risk coverage.
Lawmakers said that given the current threat landscape, allowing a lapse in terrorism coverage would expose policy holders to unnecessary financial risk and uncertainty.

Smith’s legislation would extend TRIA by seven years. The current legislation expires on Dec. 31, 2027.

The Terrorism Risk Insurance Act is also backed by the NFL, Major League Baseball, the National Association of Home Builders, the National Multifamily Housing Council, the American Council of Engineering Companies, and others.