368 Neobanks Serve 1.46B Users, Yet Two-Thirds Lack Full Licenses

Woofun AI reports that a comprehensive audit of the global neobank sector, conducted by Francesco Andreoli, Head of Developer Relations at Consensys & MetaMask, and compiled by Jiahua from ChainCatcher, has exposed a critical disconnect between market perception and operational reality. While industry analysts and venture capitalists have historically focused on funding rounds and growth metrics, this investigation, tracked via neobankbeat.com, shifts the lens to survival rates and structural integrity, revealing that the true measure of the industry’s health lies not in its entries but in its silent exits. The data underscores a sector where visibility is skewed toward launches, while the graveyard of failed entities remains largely unrecorded, creating a distorted view of risk for investors and users alike.

The scale of the current market is substantial, with 368 verified operational neobanks identified as of July 2026. These entities collectively serve approximately 1.46 billion people, a figure derived from actual customer counts reported by each company rather than theoretical total addressable market projections. The geographical distribution of these users defies traditional Western-centric fintech narratives, with Asia accounting for 817 million users. WeBank alone serves over 400 million people, a number that exceeds the combined user base of all neobanks in the United States and Europe. In Latin America, Nubank has amassed 131 million customers, surpassing the total neobank user count in the US.

Meanwhile, Europe’s leading player, Revolut, has over 50 million users, an impressive achievement that nonetheless represents only a fraction of the Asian market’s volume.

Structurally, the industry is undergoing a significant transformation driven by the emergence of web3 technologies. Among neobanks established in the 2020s that remain operational, 30% are web3 native self-custody applications, where users retain control of their balances without intermediary custody. This stands in stark contrast to the 2010s cohort, where only 4% of entities adopted this model. The broader ecosystem comprises three distinct waves: 254 traditional challenger banks, 58 fiat and crypto hybrid applications, and 56 web3 native applications. This diverse landscape is supported by 106 infrastructure providers and backed by 219 investors, indicating a complex web of dependencies that extends far beyond the consumer-facing brands. The shift toward self-custody suggests that builders are increasingly prioritizing user sovereignty, regardless of broader market sentiment toward cryptocurrencies.

A critical vulnerability within this ecosystem is the licensing deficit, which poses a fundamental risk to consumer protection. Of the 368 operational neobanks, only 127 hold a full banking license. This means that two-thirds of the entities marketed as 'banks' in mobile app stores are not legally banks. Their operational legitimacy is often rented from holding banks, electronic money license holders, or obscure card issuers. Customers are frequently unaware of this arrangement, assuming a level of regulatory oversight and deposit insurance that does not exist. This structural opacity creates a fragile foundation, where the stability of a neobank is contingent on the health of its upstream partners rather than its own financial resilience. The lack of direct licensing exposes users to counterparty risks that are rarely disclosed in marketing materials.

Historical precedents illustrate the severe consequences of this infrastructure dependency. In 2018, WaveCrest’s collapse led Visa to revoke the qualifications of a card issuer, causing dozens of crypto card projects to fail overnight. In 2020, Wirecard’s revelation of a €1.9 billion accounting hole froze funds for numerous European 'banks' that relied on its payment processing services, demonstrating how a single point of failure can cascade through the ecosystem. More recently, in 2024, the bankruptcy of Synapse, a BaaS intermediary, exposed ordinary Americans to the reality that 'FDIC insured' labels could be misleading when the underlying ledger recording ownership was compromised. These incidents highlight that when infrastructure fails, customers are left with little recourse, often receiving only a queue number in bankruptcy proceedings rather than immediate compensation.

Woofun AI data shows that the pattern of silent exits continues to this day, with recent data showing that five entities disappeared from the operational list in 2026. These companies were either liquidated, absorbed by larger players, or quietly transformed into different business models without public announcement. Unlike the high-profile collapse of FTX, which garnered extensive media coverage, these neobank failures occur without press releases or retrospectives. Apps simply cease updates, customer service becomes unresponsive, and domains redirect to partner landing pages, leaving hundreds of thousands of customers to migrate or lose their funds. This lack of visibility means that failure data is underreported, and new founders often repeat the same mistakes, unaware of the traps that have already claimed their predecessors.

The industry’s reliance on artificial intelligence as a solution to profitability challenges is also subject to scrutiny. While AI is a ubiquitous feature in neobank business plans, a rigorous review of the 368 entities reveals that only 67, or 18%, have successfully implemented operational AI models. The remaining 300+ companies are either in the pilot phase, 'exploring' potential use cases, or claiming credit for their partners' AI capabilities. This gap between hype and reality suggests that AI is not yet a mature differentiator for most neobanks. The few that have passed the test demonstrate that effective AI integration requires more than marketing claims; it demands robust data infrastructure and clear operational value, which many companies have yet to achieve.

Contrary to expectations, the most effective AI implementations are emerging from markets in the Global South, including Nigeria, the Philippines, Mexico, and Bangladesh. In these regions, traditional credit systems are often non-existent, making AI-driven credit scoring essential for serving credit invisibles. For lenders in these areas, AI is not a luxury feature but a necessity for business viability. This contrasts with Western markets, where AI is often framed as an enhancement to existing services. The success of AI in emerging markets underscores the importance of context-specific solutions, where technology addresses fundamental gaps in financial inclusion rather than optimizing already mature processes.

The concentration of infrastructure risk remains a silent killer for the industry. With 106 providers supporting 368 consumer-facing brands, the load-bearing layer of the neobank ecosystem is highly centralized. A few holding banks, BaaS platforms, and card processing companies support dozens of logos above them, creating single points of failure that are invisible to consumers. This concentration means that the failure of a single provider, such as Synapse, can impact up to 1.5 billion people indirectly. Two-thirds of neobanks are unable to survive a downturn in their upstream partners, highlighting the fragility of the current model. The lack of diversification in infrastructure providers exacerbates systemic risk, making the entire sector vulnerable to shocks that originate outside the consumer-facing layer.

Looking ahead, the licensing gap is expected to narrow as strong unlicensed players acquire or obtain licenses, while weaker entities face deletion in 2027. The first AI credit explosion is predicted to occur within two credit cycles, as the 67 operational models encounter real economic downturns that test their training data.

Additionally, the next wave of banking customers will likely be AI agents, with only 7 companies currently building infrastructure for wallets operated by agents, cards issued by agents, and machine-to-machine payments. This emerging segment mirrors the early days of web3 native applications in 2021, characterized by small, strange, but structurally significant innovations that could redefine the future of finance.