2025 Review

If you zoom out, 2025 was not a year where the industry just “added features.” It was a year where the center of gravity moved. For more than a decade, digital assets lived in a split reality with a front stage of narratives (prices, hype cycles, adoption promises) and a back stage of engineering (custody, signing, compliance, monitoring, settlement). Last year, that back stage increasingly became the product.

Stablecoins continued their shift from instruments to rails. Tokenization moved further from proof of concept into production environments. Regulators and auditors raised the bar, not by introducing new ideas, but by enforcing existing ones with greater rigor. Security incidents reminded everyone, repeatedly, that best practices are not a security model. And most importantly, institutions stopped asking whether they should “touch blockchain” and started asking what it would take to operate it like finance with controls, evidence, accountability, and reliability.

That shift is exactly where Dfns sits. It’s also why 2025 felt less like a sprint toward new logos and more like a compounding year. Each increment, a new chain, a new compliance primitive, a new security boundary, a new operating model, stacked into something broader than individual features. Below is what we achieved in 2025 and what those achievements signal about where the market is heading.

The numbers that mattered (and why they matter)

Let’s start with the scoreboard. In 2025, Dfns reached:

• $12.5M in annual revenue, a 6.25x increase compared to 2024
• $26B worth of digital assets processed through the platform
• 324 clients, including Circle, Broadridge, MoonPay, Vermont State Bank, ADI Foundation, Gemini, Republic, Paxos, and many others, a growth of more than 200% year over year
• 14+ hires, bringing the team to 35 people
• 22 blockchains integrated during the year
• 70+ product features shipped with full API and UI parity
• Selected by Sifted among the top 100 fastest growing startups in Europe

Our numbers show traction, and more importantly, convergence. The market is consolidating around wallet infrastructure as a category, and treating it less like “custody tech” and more like an operating system to architecture around and deploy from. One that orchestrates transaction workflows, authorization logic, key management, compliance integrations, audit evidence, and network coverage as programmable infrastructure rather than ad hoc tooling.

From wallet software to financial infrastructure

On the surface, scale was the most visible signal. For example, we integrated 22 blockchains over the year. But chain count alone is no longer a differentiator. Simple coverage by indexing native tokens is now table stakes. Many chains are becoming real financial networks with privacy controls, identity models, permissioning, and compliance built directly into the protocol. Supporting these networks is harder and has challenged us in our ability to adapt to fundamentally different ways of operating.

Canton became the clearest expression of this shift. In 2025, we became the first wallet infrastructure provider to deliver deep support for the Canton network, including full indexing, webhooks, and end-to-end wallet functionality. More importantly, we aligned the product with how institutions are expected to operate on Canton. We shipped Bring Your Own Validator (BYOV), allowing clients to connect their own validator nodes to Dfns and run within their own infrastructure, security assumptions, and governance boundaries. We delivered full support for Canton Coin and the CIP-56 token standard, including real time indexing, transfers, ownership states, approvals, and allowances.

This sequence matters because institutional adoption means supporting the operating model behind the chain, i.e., validator control, namespaces, privacy boundaries, reporting expectations, and workflow ergonomics. Canton is an early signal of where regulated networks are heading, and it demands wallet infrastructure that behaves like enterprise middleware rather than a freestyle product.

Velocity without fragmentation

In 2025, we also shipped more than 70 features. We enforced strict API and UI parity across the platform. Every operation available in the UI is available programmatically through APIs, and every API capability has a visible, auditable representation in the UI. 

If something exists in the UI but not in the API, manual processes creep in, and manual processes become a risk. If something exists in the API but not in the UI, behavior becomes invisible, and invisible behavior becomes an audit problem. Once parity breaks, teams invent workarounds, and workarounds inevitably turn into shadow systems.

Several product primitives shipped in 2025 were explicitly designed to reduce fragmentation:

• Multichain Wallets allow a single key to operate across multiple networks, reducing address sprawl and simplifying identity and governance. 
• Fee Sponsors introduced institutional grade fee abstraction without sacrificing policy control or auditability. 
• Audit Logs gave clients a unified, searchable record of actions across wallets and policies to improve traceability and reduce silos in investigations.
• Address Books provided a shared directory of addresses that simplifies reuse and avoids duplication across wallets and integrations. 
• Security and identity upgrades, including OIDC and SSO support, tightened the link between enterprise identity systems and transaction authorization. 

Less theater, more truth in security

Security work in 2025 followed the same pattern of uncomfortable maturation. Across the industry, incidents and disclosures forced a reality check, pushing teams away from security theater and toward real security mechanics.

At Dfns, that meant grounding security in concrete properties rather than claims. We formalized that shift by becoming certified ISO 27001, 27017, and 27018, with independent audits conducted by KPMG and MSECB. Those certifications were not badges for a website. They forced discipline into how we design systems, handle risk, manage access, document controls, and produce evidence.

One practical expression of that mindset was our native integration with Thales Hardware Security Modules (HSMs), allowing institutions to run Dfns’ orchestration stack on their own certified hardware without sacrificing performance or functionality. This matters because institutions do not buy “secure key management” in the abstract. They buy systems that fit real procurement constraints, existing vendor relationships, compliance programs, and accountability models.

Another expression was how we treated vulnerability disclosure as part of the product lifecycle itself. We published a detailed analysis of CGGMP21 vulnerabilities, including mitigation guidance, acknowledging explicitly that cryptographic protocols are not “set and forget.” We invested in broader security literacy through deep analytical writing on systemic failure modes, such as our breakdown of the Bybit/Safe incident.

In institutional environments, security is not a marketing claim. It is a continuously testable set of properties backed by evidence, controls, and operational discipline. That is what we built in 2025.

The wallet becomes a distribution layer

For years, the wallet was treated as a security boundary, a place where keys are stored and transactions are signed. Last year, that definition collapsed. What emerged instead is the wallet as a distribution layer for financial operations, where identity, compliance, authorization, liquidity, and execution converge into a single programmable surface.

Institutions are no longer asking whether they can custody assets. They are asking whether they can run financial workflows on top of them: payments, treasury management, liquidity provisioning, regulatory reporting, internal approvals, and counterparty coordination under a coherent governance model.

Compliance stopped being something you integrate and became something you design around. Identity, policy enforcement, approval logic, evidence generation, and exception handling are not add ons. They are structural elements of the transaction lifecycle.

At Dfns, this evolution materialized through how we approached third-party service integrations. E.g.:

• Our Travel Rule integration with Notabene is built directly into transaction orchestration. Notabene Flow extended that logic by addressing coordination between counterparties before funds move.
• On the yield and treasury side, Allocations with M0 allow institutions to deploy idle assets without fragmenting their stack. 
• With Swaps, including integrations with Uniswap, execution becomes just another governed operation, subject to the same controls and evidence as any other financial action.

Why the IBM partnership is groundbreaking

Becoming an official IBM technology partner is, obviously, a major milestone. But the value is not the logo. The real value is what IBM validates about the market and what the partnership unlocks for institutions that want to enter digital assets without breaking their existing risk and compliance models.

In October 2025, IBM launched IBM Digital Asset Haven with Dfns as a core partner and positioned it for banks, governments, and regulated enterprises. The message was clear: digital assets are no longer an experimental layer; they’re becoming part of the core financial stack.

IBM brings something few companies can: global legitimacy. For compliance, legal, risk, and procurement teams, IBM is already an approved vendor. That changes everything. Digital asset infrastructure no longer needs to enter regulated organizations as an exception. It’s pre-accepted through the same procurement channels, governance frameworks, and control processes that already exist. This alone removes one of the biggest barriers to adoption.

IBM is also a powerful distribution engine. Its technology supports more than 70% of the world’s financial transactions. Its teams are embedded inside the largest banks, market infrastructures, payment networks, and governments. When IBM introduces a digital asset platform, it is not positioned as a niche product. It is positioned as infrastructure.

What makes the partnership truly important is the depth of the integration. It’s about building security and operating models that match how institutions actually run their systems:

• IBM Hyper Protect Virtual Servers (HPVS) for confidential computing and hardened runtimes
• IBM Offline Signing Orchestrator (OSO) for air-gapped cold signing without manual processes
• HSM-based signing tightly integrated with IBM infrastructure controls
• IBM Z and LinuxONE for deployment on the same mainframe-class systems that run core banking and critical financial workloads

This is now the recurring pattern we see. Institutions want digital asset capability, but they want it inside their existing security perimeter. If a platform cannot integrate with HSMs, confidential computing, cold-signing workflows, IAM, IDP, and compliance tooling, it’s not institutional-grade and must be surrounded with extra controls. In 2025, we chose the opposite approach. We built the platform so these controls are native and not added later. And by doing it with IBM, we anchored that model inside the same systems, vendors, and trust frameworks that already run the global financial system.

Not just growing, compounding

In 2025, we invested deliberately in foundational literature for the industry; explaining real threats, comparing architectures with intellectual honesty, and drawing a clear line between what merely works in a demo and what actually survives in production.

That meant publishing deep technical comparisons, such as our architectural analysis of Dfns vs Fireblocks across more than 420 API endpoints, framed not as a feature checklist but as a discussion of system design. It meant developing a rigorous security narrative, grounded in real-world incident analysis and protocol-level vulnerability disclosure. And it meant writing technical product narratives that explain not just what exists, but why certain primitives are necessary if this industry is to mature, like the most comprehensive “Wallet Service Guide” in the industry (21-minute read).

The industry lacks a shared language for wallet infrastructure. Without that language, every team keeps rebuilding the same fragile patterns, making the same mistakes, and rediscovering the same constraints. Good infrastructure compounds not only through code, but through understanding.

This philosophy of compounding also shaped how we thought about growth. Being selected by Sifted among the top 100 fastest-growing startups in Europe was gratifying, but it was never the real metric. The more important question was always: what kind of growth is it? In infrastructure, the healthiest growth can be measured through expansion inside existing customers as they scale. Or adoption by conservative institutions that care deeply about auditability and control. It’s partnerships that demand real integration depth, not just logo alignment. It’s product surface area that reduces vendor sprawl instead of increasing it. That is the growth profile 2025 consistently reinforced.

If 2025 was the year wallet infrastructure became financial infrastructure, then 2026 will be about who can deliver that infrastructure with stronger evidence and audit primitives, more composable compliance, deeper integration into enterprise security perimeters, and a broader set of money operations—execution, treasury, settlement—governed by a unified and coherent OS.

Our long-term bet remains unchanged. The winners in onchain finance will be the teams that turn blockchain operations into something institutions can run with the same confidence they already apply to payments, custody, and settlement. In 2025, Dfns made that bet tangible in revenue, in clients, in integrations, in security accomplishments, in compliance primitives, in platform partnerships, and in the quiet discipline of shipping a coherent system. That is the compounding we intend to keep pushing.

Get started today: dfns.co