a16z: 17 New Crypto Directions to Excite Us in 2026

Edited by: Sonal Chokshi, a16z

Compiled by: Tim, PANews

Stablecoins, RWA, Payments and Finance

A Better, Smarter Stablecoin On/Off-Ramp

Last year, the trading volume of stablecoins was estimated at $4.6 trillion, continually hitting new record highs. Specifically, this is more than 20 times the transaction volume of PayPal; nearly three times that of Visa, one of the world’s largest payment networks; and quickly approaching the transaction scale of the US Automated Clearing House (ACH), which handles direct deposits and other financial transactions.

Today, stablecoin transfers can be completed in under 1 second at a cost of less than one cent. However, the unresolved issue is how to connect these cryptocurrencies with the financial infrastructure people use daily. In other words, establishing exchange channels between stablecoins and traditional currencies.

New generation startups are filling this gap, linking stablecoins with mainstream payment systems and local currencies. Some use crypto verification technology to allow people to exchange local account balances for digital dollars. Others connect regional payment networks, utilizing QR codes, real-time payment systems, and other features to enable interbank transfers. Still, some are building truly interoperable global digital wallet layers and card issuance platforms, allowing users to pay with stablecoins at everyday merchants. These innovations broaden participation in the digital dollar economy and are expected to accelerate stablecoins becoming a mainstream payment method.

As these on/off-ramps mature, digital dollars will directly integrate with local payment systems and merchant tools, spawning new behavioral patterns. Workers can receive cross-border salaries in real time, merchants can accept globally circulating digital dollars without bank accounts, and payment apps can settle value instantly with users worldwide. Stablecoins will fundamentally transform from fringe financial tools into the internet’s basic settlement layer.

— Jeremy Zhang, a16z crypto engineering team

Understanding RWA and Stablecoins in a More Native Way

We observe that banks, fintech companies, and asset managers are highly interested in bringing traditional assets like US stocks, commodities, and indexes onto the blockchain. As more assets are tokenized, their tokenization often remains superficial, limited to the current concept of real-world assets, without fully leveraging crypto-native features.

However, synthetic products like perpetual contracts can provide deeper liquidity and are often easier to implement. Perpetual contracts also offer understandable leverage mechanisms, making them, I believe, the most product-market-fit crypto-native derivatives. I also think emerging market equities are among the most suitable asset classes for perpetualization. (In some cases, the liquidity of zero-date options markets on certain stocks even exceeds that of the spot markets, creating compelling experimental cases for perpetualization.)

Ultimately, it’s a question of “perpetualization versus tokenization.” But regardless, we expect to see more crypto-native RWA tokenization within the next year.

Along similar lines, by 2026, we will see more “native issuance” of stablecoins, not just tokenization. Stablecoins have become mainstream in 2025, with the number issued continuing to grow.

However, stablecoins lacking robust credit infrastructure resemble narrow banks, holding specific highly liquid assets deemed ultra-safe. While narrow banks are legitimate financial products, I believe they will not be the backbone of the on-chain economy in the long run.

Recently, many new asset managers, curators, and protocols have started offering asset-backed loans on chain, collateralized by off-chain assets. These loans are usually initiated off-chain and then tokenized. I think tokenization offers little benefit here except for distribution to on-chain users. That’s why debt assets should be issued directly on-chain, not tokenized after off-chain issuance. On-chain issuance reduces loan management and backend costs and improves accessibility. The challenging parts will be compliance and standardization, but builders are already working on these issues.

— Guy Wuollet, general partner at a16z crypto

Stablecoins Kickstart Banking Ledger Upgrades and New Payment Scenarios

Banking software systems are often unfamiliar to modern developers: in the 1960s-70s, banking was a pioneer in large-scale software systems. The second-generation core banking systems emerged in the 1980s-90s (e.g., Temenos GLOBUS, Infosys Finacle). But these systems are aging and update speeds are slow. As a result, key core ledgers recording deposits, collateral, and other debts still often run on mainframes, programmed in COBOL, interacting via batch files rather than APIs.

Most of the world’s assets depend on these decades-old core ledgers. While these systems are proven, trusted by regulators, and deeply integrated into complex banking scenarios, they also hinder innovation. Adding real-time payment features can take months or years, requiring overcoming layers of technical debt and regulatory complexity.

This is where stablecoins matter. The past few years have been about stablecoins finding product-market fit and entering the mainstream. This year, traditional financial institutions have embraced them more than ever. Stablecoins, tokenized deposits, tokenized bonds, and on-chain bonds enable banks, fintechs, and financial institutions to develop new products and serve new clients. More importantly, they do so without forcing these institutions to rewrite their long-established, stable legacy systems. Thus, stablecoins offer a new path for institutional innovation.

— Sam Broner

Web3 Banking Goes Mainstream

As intelligent agents proliferate and more business activities happen automatically in the background rather than via user clicks, the way value flows from money must change.

In a world driven by intent rather than step-by-step instructions, AI agents recognizing needs, fulfilling obligations, or triggering outcomes can mobilize funds, and value must flow as quickly and freely as information does today. This is where blockchain, smart contracts, and on-chain protocols come into play.

Smart contracts can now settle global USD payments in seconds. By 2026, emerging primitives like x402 will make settlements programmable and responsive: no invoices, no reconciliation, no batch processing—immediate, permissionless payments of data, GPU compute, or API calls between agents. Developers’ software updates will carry built-in payment rules, limits, and audit trails, all without fiat integration, merchant onboarding, or financial institutions. Markets will self-settle in real time as events unfold, with odds dynamically updating, autonomous agents trading freely, and global payouts happening in seconds—all without custodians or exchanges.

Once value can flow this way, “payment flows” cease to be a separate layer and become a network behavior: banks become the internet’s backbone, and assets become infrastructure. When money turns into internet-routable information packets, the internet is no longer just the backbone of finance; it becomes finance itself.

— Christian Crowley and Pyrs Carvolth, a16z crypto GTM team

Democratization of Wealth Management

Traditionally, personalized wealth management has been exclusive to high-net-worth clients: offering tailored advice and customized portfolios across asset classes is costly and complex. But with tokenization of more assets, combined with AI-driven recommendations and collaborative strategies, personalized management can be executed and rebalanced instantly and at low cost.

This is not just about robo-advisors; everyone can access active portfolio management, not just passive. By 2025, traditional financial institutions will have increased their exposure to crypto (via direct investments or ETPs), but that’s just the beginning. By 2026, platforms dedicated to “wealth growth,” not just “wealth preservation,” will emerge. Fintech companies like Revolut and Robinhood, and centralized exchanges like Coinbase, will leverage their tech advantage to capture more market share.

Meanwhile, DeFi tools like Morpho Vaults can automatically allocate assets into the best risk-adjusted yield lending markets, providing core yield assets for portfolios. Holding residual liquidity in stablecoins rather than fiat, and investing in RWA money market funds instead of traditional ones, can further boost yields.

Finally, retail investors can more easily invest in illiquid private markets—such as pre-IPO companies and private equity—via tokenization, unlocking new potential while maintaining compliance and reporting standards. As various assets in balanced portfolios are tokenized (bonds, stocks, private and alternative investments), rebalancing can happen automatically without fund transfers.

— Maggie Hsu, a16z crypto GTM team

AI and Agents

From “Know Your Customer” (KYC) to “Know Your Agent” (KYA)

The constraints on intelligent agent economies are shifting from intelligence to identity verification.

In finance, “non-human identities” outnumber human employees by 96 times, yet these identities remain unaccounted for—ghost accounts. The key missing foundation is KYA: Know Your Agent.

Just as humans need credit scores to get loans, AI agents need cryptographically signed credentials to transact, linking the agent to its authorized entity, operational limits, and accountability. Until this mechanism matures, merchants will continue to block agents at firewalls. Decades of KYC infrastructure must now rapidly address KYA in months.

— Sean Neville, co-founder of Circle, architect of USDC, CEO of Catena Labs

We Will Use AI to Complete Research Tasks

As a mathematical economist, I found it hard in January to get general AI models to understand my workflows. By November, I could give the models abstract instructions as if guiding a PhD student, and they sometimes produce novel and correct answers. Beyond my personal experience, we are witnessing AI being applied more broadly in research, especially in reasoning: current models can directly assist scientific discovery and autonomously solve Putnam math competition questions (arguably the most difficult university-level math exam globally).

Open questions remain about which research domains benefit most from such tools and how they will be used. I predict AI research will give rise to and reward a new scholarly style: one that favors the ability to infer relationships between concepts and quickly deduce from more speculative answers. These answers may not be precise but can still point in the right direction (at least within a certain topology). Ironically, it’s like harnessing hallucination: when models are “smart” enough, giving them abstract, divergent thinking space can produce nonsensical content, but sometimes breakthroughs happen—similar to human creativity in nonlinear, uncertain explorations.

This kind of reasoning will require a new AI workflow, involving nested models that help researchers evaluate early ideas, filter out falsehoods, and distill valuable insights. I’ve used this approach for paper writing, patent searches, creating new art, or—unfortunately—discovering new smart contract attacks.

However, running such nested-agent research systems demands better interoperability among models and mechanisms to recognize and reasonably compensate each contribution—areas where encryption tech could help.

— Scott Kominers, a16z crypto researcher, Harvard Business School professor

Invisible Tax on the Open Web

The rise of AI agents is imposing an invisible tax on the open web, fundamentally disrupting its economic basis. This stems from the growing mismatch between the contextual layer (the web’s environment) and the execution layer: today, AI agents extract data from ad-supported sites (context layer), providing convenience while systematically bypassing revenue channels that support content creation (ads, subscriptions).

To prevent erosion of the open web and preserve diverse content creation fueling AI, we need large-scale technical and economic solutions—new sponsorship models, attribution systems, or other funding mechanisms. Existing AI licensing agreements are stopgap measures, often compensating content creators only a fraction of lost traffic revenue.

The web needs a new economic model where value flows automatically. Over the next year, a key shift will be from static licensing to real-time, usage-based compensation—testing and deploying such systems, possibly enabled by blockchain-powered micro-payments and precise provenance standards, to automatically reward entities providing information enabling successful AI agent tasks.

— Liz Harkavy, a16z crypto investment team

Privacy and Security

Privacy Will Become the Most Critical Moat in Crypto

Privacy is a key requirement for global on-chain finance, yet it’s lacking in nearly all existing blockchains today. For most blockchains, privacy features are an afterthought or supplementary.

But now, privacy itself can distinguish a blockchain from all others. It also plays a more vital role: creating on-chain lock-in effects—what we might call “privacy network effects.” Especially in a world where performance alone can no longer guarantee competitive advantage, this is crucial.

With bridging protocols, if all information is public, migrating assets across chains is easy. But once private info is involved, the situation changes: bridging tokens is simple, but bridging secrets is very hard. When entering or leaving private zones, there’s always a risk that observers on monitored chains, mempools, or network traffic can de-anonymize users. Crossing between private and public chains, or between two private chains, leaks metadata like timing and transaction size correlations, making tracking easier.

Compared to many homogeneous new chains (whose fees might plummet due to competition, as block space is no longer a differentiator), privacy-enabled chains can generate stronger network effects. The reality is, if a “general-purpose” public chain lacks a thriving ecosystem, killer apps, or distribution advantages, users and developers have little reason to use or build on it, let alone stay loyal.

When users transact on public chains, switching between chains is easy. But when using private chains, choice of chain is critical: once onboarded, migration is unlikely, and privacy exposure is a risk—creating winner-takes-all dynamics. Since privacy protection is crucial for most real-world use cases, a few privacy-preserving chains could dominate the entire crypto market.

— Ali Yahya, a16z crypto general partner

The Future of Messaging Will Need to Be Quantum-Resistant and Decentralized

As the world prepares for the quantum era, many cryptography-based communication apps (e.g., Apple iMessage, Signal, WhatsApp) have led the way. But all mainstream messaging relies on trust in private servers operated by a single organization—targets for government shutdowns, backdoors, or coercion.

If a country can shut down your personal servers, if a company holds the keys to your private servers, or even if a company owns the servers, what’s the point of quantum encryption? Private servers depend on “trust me,” but without private servers, “you don’t need to trust me.” Communication shouldn’t require a middleman. It needs open protocols—trustless, peer-to-peer, and with top-tier encryption, resistant to quantum threats. In an open network, no individual, company, nonprofit, or government can deprive us of communication. Even if one country or company shuts down an app, hundreds of new versions will emerge overnight. Even if a node is shut down, economic incentives via blockchain tech will bring new nodes immediately.

When people can own their information via private keys as they do money, everything changes. Apps may come and go, but users will always control their data and identities—gaining true ownership without relying on a particular app.

It’s not just about quantum resistance or cryptography; it’s about ownership and decentralization. Without both, what we build is just a seemingly unbreakable encryption system that can still be shut down anytime.

— Shane Mac, XMTP Labs co-founder and CEO

Privacy-as-a-Service

Behind every model, agent, and automation flow is a simple element: data. Today, most data pipelines—inputs and outputs—are opaque, volatile, and hard to audit. This can work for some consumer apps or permissioned systems, but many industries (like finance and healthcare) require protecting sensitive data privacy. It’s also a major obstacle for institutions aiming to tokenize RWAs.

How can we enable innovation that is secure, compliant, autonomous, and globally interoperable while protecting privacy? There are many ways, but I want to focus on access control: who controls sensitive data? How does it flow? Who (or what) can access it?

Without data access controls, users wanting confidentiality rely on centralized services or custom systems—time-consuming, costly, and limiting the potential of traditional institutions to unleash on-chain data management. As intelligent agents start autonomously browsing, trading, and making decisions, cross-industry users and institutions need cryptographic verification rather than reliance on “trust me” models.

This is why I believe in “Privacy-as-a-Service”: new tech that offers programmable native data access rules, client-side encryption, and decentralized key management—precise control over who can decrypt what, when, and under which conditions, all on-chain. Coupled with verifiable data systems, privacy protection becomes a core part of internet infrastructure—not just an afterthought patch—making privacy a foundational element.

— Adeniyi Abiodun, co-founder and chief product officer at Mysten Labs

From “Code Is Law” to “Rules Are Law”

Recently, some battle-tested DeFi protocols faced hacks, despite strong teams, rigorous audits, and years of stable operation. These incidents reveal an unsettling reality: current security standards rely heavily on case-by-case judgment and experience.

To mature, DeFi security must shift from vulnerability detection to design principles, from “do your best” to “principled” approaches:

In static deployment, pre-launch phases (testing, auditing, formal verification), this means systematically verifying global invariants rather than just partial ones manually checked. Many teams are developing AI-assisted proof tools to help write technical specs, propose invariants, and significantly reduce the cost of formal verification.

In dynamic, post-deployment phases (runtime monitoring, enforcement), these invariants can become dynamic fences—last lines of defense—encoded as runtime assertions that each transaction must satisfy.

Thus, we no longer assume all vulnerabilities can be found; instead, we enforce key security properties directly in code, with violations leading to automatic reversion.

This is not just theoretical. In practice, nearly all exploit attempts trigger one of these security checks, potentially thwarting malicious attacks. The once-popular “code is law” mantra is evolving into “rules are law”: even new attack vectors must satisfy the security properties that preserve system integrity. As a result, the feasible attack surface shrinks, or at least becomes extremely difficult to execute.

— Daejun Park, a16z crypto engineer

Other Sectors and Applications

Prediction Markets Will Be Larger, Broader, Smarter

Prediction markets are becoming mainstream. Next year, with integration with crypto and AI, they will be bigger, broader, smarter—but also pose new challenges for entrepreneurs.

First, more contracts will be listed. This means we can get real-time odds for major elections or geopolitical events, as well as obscure outcomes and complex intersections. As these new contracts emerge, providing more information and becoming part of the news ecosystem (already happening), they will raise societal questions: How to value this info? How to design markets to be more transparent, auditable, and versatile? All achievable via crypto.

To handle the surge of contracts, we need new consensus mechanisms to verify authenticity. Centralized verdicts (e.g., did the event happen? How to confirm?) are crucial but limited—as shown by controversies like Zelensky lawsuits or Venezuela elections. To resolve fringe cases and expand prediction markets’ practical uses, decentralized governance and large language model oracles will help determine facts amidst disputes.

AI’s predictive power is already impressive. For example, AI agents operating on these platforms can scan global signals, gaining short-term trading advantages—helping us explore new cognitive dimensions and improve future predictions. These agents can serve as advanced political analysts, revealing factors behind social events.

Can prediction markets replace polls? No, but they can improve them—and poll data can feed into prediction markets. As a political scientist, I’m interested in how prediction markets can cooperate with vibrant polling ecosystems. We need new tech like AI and crypto to improve survey experiences, verify respondents as real humans, and create new ways for them to participate.

— Andy Hall, a16z crypto research advisor, Stanford political economy professor

The Rise of Betting-Driven Media

Objectivity has been fractured in traditional media for some time. The internet gives everyone a voice. Now, more operators, practitioners, and builders directly speak to the public. Their views reflect their interests, and contrary to intuition, audiences respect them. They welcome their vested interests.

The innovation is not just social media’s rise but the arrival of cryptography tools enabling public, verifiable commitments. AI makes cheap, unlimited content generation possible—any viewpoint, any identity, real or fictional. Relying solely on human or bot statements is no longer enough. Tokenized assets, programmable escrow, prediction markets, and on-chain records provide a stronger trust foundation: a commentator can publish an argument and prove they are genuinely putting their money where their mouth is. A podcaster can lock tokens as a sign they’re not opportunistically trading or “pump and dump.” An analyst can link predictions to publicly settled markets, creating an auditable trail.

This is the early form of “betting media”: a media that recognizes “vested interests” and can prove them. Trust no longer comes from pretending to be neutral or just claiming; it derives from actual stakes—verifiable commitments to real bets. Betting media will not replace other forms but complement them. It signals a new trust paradigm: not “trust me, I’m neutral,” but “here is the risk I’m taking, verified by what I’m staking.”

— Robert Hackett, co-editor of a16z crypto

Cryptocurrency as a New Building Block, Beyond Blockchain

For years, SNARKs (cryptographic proofs that verify computations without recomputation) have been mainly used in blockchain. Their overhead is huge: generating proofs can be a million times more work than the computation itself. This is worthwhile when verified by thousands of nodes but impractical elsewhere.

This will change. By 2026, zkVM provers’ overhead will drop to about ten thousand times, with only hundreds of megabytes of memory needed, fast enough for smartphones, and costs low enough to deploy everywhere. The key is that GPU parallel throughput is about ten thousand times that of a laptop CPU, enabling real-time proof generation by single GPUs.

This could unlock visions from old research papers: verifiable cloud computing. If your workload runs in the cloud, whether due to insufficient scale for GPU or lack of expertise, you will be able to obtain cryptographic proofs of correctness at a reasonable cost. These proofs will be optimized for GPUs, requiring no changes to your code.

— Justin Thaler, a16z crypto researcher, associate professor at Georgetown CS

Lightweight Trading, Heavy Construction

Treating trading as a transit point rather than a destination is the business model of crypto companies.

Today, aside from stablecoins and some core infrastructure, every thriving crypto company seems to be shifting or planning to shift into trading. But what if “every crypto company becomes a trading platform”? The industry would look different. Many firms doing the same thing lead to a race to the bottom—only a few winners remain. This means those rushing into trading may miss the opportunity to build more resilient, durable business models.

While I sympathize with founders trying to keep their companies afloat, chasing immediate product-market fit has costs. In crypto, this is especially true. The industry’s atmosphere around tokens and speculation often tempts founders to seek instant gratification rather than product-market fit—like a cotton candy experiment.

Trading itself is fine; it’s an important market function. But it doesn’t have to be the end. Those focusing on the “product” side of PMF might have a better shot at being the final winners.

— Arianna Simpson, general partner at a16z crypto

Regulatory-Technology Alignment to Unlock Blockchain’s Full Potential

Over the past decade, one of the biggest barriers to creating blockchain in the US has been legal uncertainty. Securities laws have been misused and selectively enforced, forcing founders to follow regulations designed for traditional companies, not blockchain. Over the years, firms replaced product strategy with legal risk minimization, and engineers took a backseat while lawyers took center stage.

This has led to strange phenomena: founders advised to stay opaque; token distribution made ad hoc to dodge laws; governance reduced to show; organizational structures driven by compliance, not efficacy; token designs that deliberately obscure economic value or avoid business models. Worse, projects operating at regulatory edges often outperform honest builders.

But the market is now closer than ever to legal clarity through regulation, which could eliminate these distortions next year. If legislation passes, it will incentivize transparency, establish clear standards, and create structured pathways for fundraising, token issuance, and decentralization—replacing the current “regulation roulette.” Since the GENIUS law passed, stablecoins have exploded; broader legislation around market structure could be even more transformative, focusing on the network ecosystem.

In essence, such regulation will enable blockchain to truly operate as a network: open, autonomous, composable, neutral, and decentralized.

— Miles Jennings, a16z crypto policy team and general counsel