It’s been about 15 years since the American banking industry nearly collapsed. Between 2008 and 2011, 414 chartered banks failed in the United States as a result of the Great Recession. When you include forced mergers, acquisitions, consolidations, and other exits, the real number of banks affected is much higher. Roughly 2,000 banks across the U.S. disappeared in the years after 2008, representing about one in four.

The shrinkage of the banking industry didn’t reduce Americans’ access to banking; instead, it concentrated it. The largest U.S. banks absorbed the business, branches, and assets of the regional banks hit hardest by the crisis. As their market share grew, so did their reliance on digital infrastructure to manage it. Mobile apps replaced branches, and lending decisions that once depended on local knowledge became more standardized and algorithmic.

Venture capital in Silicon Valley rebounded sharply after the crisis, and a new generation of technology companies directed that capital toward financial services, filling the gaps left by the decline in local banking. Now that the first wave of fintech disruptors is maturing, a more consequential second wave is forming behind it. Cryptocurrency and artificial intelligence are rapidly changing how money moves and how financial decisions are made, and the next era of finance will be defined by whoever builds the infrastructure they run on.

The Slow Decline of Local Banking

One of the domino effects of the financial crisis was the acceleration of banking’s shift from physical to digital.

As the industry consolidated and cost pressures increased, maintaining large branch networks became harder to justify. Between 2008 and 2020, more than 13,000 bank branches closed nationwide; the COVID-19 pandemic accelerated the trend, with another 4,000 branches closing between 2020 and 2022.

At the same time, customers were becoming more comfortable managing their finances online: digital dashboards improved, cashless payments became more common, and the infrastructure behind online financial services matured. Digital banking replaced the loss of service caused by branch closures, even expanding banking access in regions that were previously underserved.

But as the 2010s progressed, innovation in digital banking (specifically among the big banks) began to slow. Deloitte noted that many banking systems were “never optimized to take advantage of new technologies,” leaving banks with increasingly costly and inflexible infrastructure. As innovation accelerated elsewhere—particularly in crypto and modern interface design—legacy banking products started to feel increasingly outdated.

Regional banks, meanwhile, lacked the resources and technical infrastructure to build competitive digital experiences, causing some customers to drift away. What went with them was the local banking experience. Regional banks have a certain charm that gets lost in the switch to an app; however, as branches closed and online banking became the norm, the benefits of adopting modern solutions began to outweigh the family feel.

It created a bad deal for the customer. Large institutions replicated the basics of banking online without meaningfully improving them, while smaller banks offered better relationships but couldn’t keep up technologically.

Reinventing the Financial Stack

While Wall Street maintained the status quo, Silicon Valley sought to disrupt it. The success of startups like Uber and Airbnb showed that no industry was too large to resist innovation, prompting a wave of challengers to the old financial order.

There was already proven success for disruptive financial technology. PayPal was an inflection point, founded in 1998 as a response to the growth of ecommerce that coincided with the rise of Amazon, eBay, and other online marketplaces. It showed that much of the potential in fintech was in the mediation between the bank and customer: easier ways to access funds, move money, accept payments, and make financial decisions.

In 2009, Bitcoin introduced a new way to interact with money, proposing a financial system with no central authority. It was extreme and largely impractical for everyday commerce, but it signaled something important: people were open to rethinking what money could be and who controlled it. In the wake of the financial crisis, banks weren’t just lagging in innovation, but they were also losing ground in the broader debate over trust and control.

The years immediately following the 2008 crisis saw the first crop of serious fintech challengers. Each of these companies attacked a specific chokepoint in the legacy financial system, and each found a receptive audience:

Square launched in 2009, putting a card reader in the hands of small merchants who had previously been locked out of accepting credit cards.
Stripe arrived a year later, making it trivially easy for developers to accept payments online and effectively removing one of the last major technical barriers to starting an internet business.
Lending Club and Prosper pushed peer-to-peer lending into the mainstream, offering borrowers better rates than traditional banks while giving investors a new asset class.
Robinhood, founded in 2013, began dismantling the commission-based brokerage model entirely, democratizing stock trading for a generation that had grown up watching their parents lose retirement savings in the crash.
Mint and Personal Capital brought financial visibility to consumers who didn’t have access to a clear picture of where their money was going.

By the late 2010s, a second and more ambitious wave arrived, this time targeting businesses rather than consumers. The focus shifted from making banking more accessible to making it more intelligent:

Brex launched in 2017, offering corporate cards to startups that traditional banks routinely turned away for lacking the credit history or revenue that legacy underwriting models required.
Mercury followed with a banking platform designed for technology companies, offering clean interfaces and API access as an alternative to traditional business banking workflows.
Ramp came shortly after, focusing on helping companies manage and reduce spend by combining a card product with expense management and spend controls.

These platforms did more than move banking online. They changed what businesses expected from a bank by tying financial services directly into the tools they already used and by making core workflows faster and more visible.

So the stage was set: businesses and consumers wanted the smaller feel of regional banking that made them trust their financial partners, but they needed modern products that either matched or improved upon what the big banks could offer. The companies that figured out how to deliver both would define the next era of finance.

Outside of the core appeal of moving away from the big banks, the next frontier of fintech is taking shape in two key areas: crypto and AI.

Crypto Finds a Practical Purpose

Although crypto has been a force in the financial world for more than a decade, the past five years have seen a large-scale normalization and adoption of blockchain technology. The biggest driver behind that adoption was the money-making possibilities of exchange-traded assets like Bitcoin and Ethereum.

The pandemic acted as rocket fuel. Bitcoin’s value multiplied tenfold from October 2020 to November 2021. The Journal of International Money and Finance found that massive liquidity injections, low interest rate policies, and fiscal stimulus heightened activity across capital markets during the pandemic. The conditions made Bitcoin's independence from central bank policy a uniquely attractive investment.

Fidelity Digital Assets argued that investors concerned about currency debasement and inflation eroding returns should consider Bitcoin a "preeminent macro asset." In an environment where the dollar was losing purchasing power, Bitcoin emerged as a new hard asset in American portfolios.

Then, things cooled. Crypto winter is the industry term for the chill that descended on the market in 2022. After the collapse of the TerraUSD algorithmic stablecoin, other major crypto players fell like dominos. The FTX implosion that November, which exposed Sam Bankman-Fried's alleged misuse of billions in customer funds, was particularly damaging. Bitcoin collapsed to around $16,000, and the total crypto market cap shed roughly two-thirds of its value from its peak.

But the underlying infrastructure kept developing through the winter. The problem that had always limited crypto's operational appeal remained: exchange-traded cryptocurrency does not make practical sense for everyday commerce. Using an asset as volatile as Bitcoin for a payment means you could wind up woefully underpaying or overpaying, so Bitcoin remained largely a speculative instrument traded on exchanges like Coinbase and Kraken.

The blockchain itself, however, offered something compelling that had nothing to do with speculation. Unlike bank-to-bank rails, which charge processing fees and can take days to settle a payment, on-chain transactions are low-fee and nearly instant, operating around the clock without the constraints of banking hours or correspondent networks.

Stablecoins emerged as the answer to the volatility problem. Pegged one-to-one to a fiat currency (typically the US dollar) and backed by liquid reserves like Treasury bills, stablecoins preserve the blockchain’s efficiency without the price swings that made Bitcoin impractical for payments.

Tether (USDT) and Circle (USDC) became the dominant issuers, and the numbers behind their adoption have been striking. The stablecoin market supply now exceeds $300 billion, up from less than $30 billion in 2020. Real-world stablecoin payment volume, meaning actual commerce rather than crypto trading activity, doubled in 2025 to $400 billion, with roughly 60 percent of that driven by B2B payments.

The regulatory environment has begun to catch up as well. Europe became the first major market to establish a comprehensive framework for stablecoins with the Markets in Crypto-Assets regulation in 2024, and the US followed with the GENIUS Act in 2025, the first federal law establishing rules for payment stablecoins.

The Three Layers of Financial AI

The debut of ChatGPT in late 2022 did for AI what Bitcoin did for crypto: it made an abstract technology suddenly legible to a mainstream audience. But with the rise of Anthropic’s Claude models and the broader economic implications of AI’s potential, it was only the beginning of a more consequential shift.

Large language models (LLMs) are the underlying infrastructure, predictive systems trained on vast amounts of data to understand and generate language. What most people experience as AI is little more than interacting with the LLM: a conversational interface that’s good at answering questions and drafting text.

Agents are different. Rather than merely responding to prompts, they can execute tasks. They can break a goal into steps, make decisions, call external tools, and complete workflows without a human in the loop at each stage. Lightweight agents like Claude’s Cowork model can make changes to files on your computer; more heavyweight setups, like OpenClaw, let an agent do almost anything you ask, from sending emails to purchasing items online, and even running in the background without you noticing.

The use case potential for AI, especially regarding what it can do inside a business, largely lies in agentic capabilities. A chatbot can tell you what your burn rate is, but an agent can do something about it.

While the role of artificial intelligence in financial workflows is still maturing, early use cases are already proving to be transformative. There are three elements of AI in finance that a mature solution needs to handle: analysis, authorization, and orchestration.

Analysis is the most straightforward. Give an AI secure access to financial metrics from a banking dashboard, and it can perform the kind of complex financial planning and analysis work that used to require a dedicated analyst.

The question of how much autonomy an AI system should have to initiate, approve, or block financial transactions is what developments in authorization have sought to answer. A system that flags suspicious spend is useful. A system that autonomously executes or blocks a payment on behalf of a business requires a different level of institutional confidence, regulatory clarity, and auditability.

Google announced the Agent Payments Protocol (AP2) in September 2025, an open-source framework that allows AI agents to search for and make purchases on a consumer's behalf using structured authorization mandates. Notably, the protocol creates a non-repudiable audit trail, meaning every transaction is permanently recorded and cannot be altered or denied after the fact. Google has since partnered with over 25 payments-related entities including American Express, PayPal, Coinbase, Mastercard, and WorldPay.

Overall, authorization guardrails are mature enough that the fintech industry has moved on to solving harder problems.

Orchestration is the final and most ambitious layer. Whereas authorization is a binary decision (should this transaction proceed, yes or no?), orchestration acts on multiple variables in a financial transaction to complete a workflow from end-to-end without human intervention. Say you prompt an AI to send an international payment to a supplier. The AI selects the right payment rail, handles compliance across jurisdictions, optimizes the routing path, and updates records across multiple systems.

The Next Financial Moment

The platforms that emerged from the fintech wave of the late 2010s, such as Ramp, Brex, and Mercury, each built something useful by attacking a specific aspect of the gaps left in the financial industry. Slash is making a different bet: that business banking will unify programmable money and intelligent automation into a single account, governed by the same controls, available to any business from day one, without requiring a team of engineers to stitch it together.¹

Slash is a business banking platform built around the premise that stablecoins and AI aren't add-on features, but the core infrastructure for modern finance.⁴

Most companies exploring stablecoins today have to maintain separate infrastructure for on-chain payment activity, creating the kind of complexity that modern financial tooling is supposed to solve. Slash uses stablecoins as a standard payment rail without needing a separate relationship with a crypto exchange, meaning your banking and crypto capabilities are in the same place.

With a Slash account, your business can receive a payment on-chain, convert it to USD using built-in on- and off-ramps, pay a contractor across the world with stablecoins, and reconcile the whole transaction without touching a separate platform. The cost and settlement advantages that make stablecoin rails compelling – low fees, near-instant finality, twenty-four-hour availability – are accessible from the same interface used to run the rest of the business.

Slash's crypto-native architecture also unlocked something previously out of reach for most international businesses: a full-featured U.S. bank account without a U.S. incorporation. The Slash Global USD Account backs assets with the USDSL stablecoin, using the blockchain to replace the correspondent banking infrastructure that made cross-border banking slow, expensive, and inaccessible.³ Business owners anywhere in the world can now access a USD account with full payment capabilities and corporate cards through the Slash platform. Stablecoins made it technically possible; Slash was the first to build it.

Slash is also the first business banking platform to develop the three-layer framework that defines what a capable financial AI product looks like. Twin, Slash's integrated AI financial agent, handles analysis, authorization, and orchestration, covering the full financial workflow from a single interface:

Analysis is the first layer. Twin can access your Slash account to answer questions about runway, burn rate, vendor spend, and much more without needing to assemble a data pipeline or exporting CSVs.
Authorization is handled through scoped permissions and spending controls that define exactly what an agent is allowed to do, with every action recorded as an auditable mandate rather than an opaque API call. This is the piece the broader industry is still struggling to get right: the difference between an AI that makes recommendations and one that can actually act, within defined guardrails, with a clear record of what it did and why.
Orchestration is where Slash has driven the most change, building workflows that span approval, payment-rail selection, execution, and bookkeeping — carried out by an AI agent operating natively within the platform. If you ask Twin to handle a new subscription, it will spin up a new virtual card; if you ask Twin to settle an invoice, it will look at past transaction history and payment terms to choose the correct rail and timing based on your cash flow.

Finance is moving beyond traditional banking. Slash gives businesses access to the infrastructure shaping what comes next.