a16z: 17 Key Potential Trends in Crypto to Watch by 2026
TechFlow Selected TechFlow Selected
a16z: 17 Key Potential Trends in Crypto to Watch by 2026
Covering agents and artificial intelligence, stablecoins, tokenization and finance, privacy and security, extending to prediction markets, SNARKs, and other applications.
Navigating Web3 tides with focused insightsWritten by: Adeniyi Abiodun, Ali Yahya, Andrew Hall, Arianna Simpson, Christian Crowley, Daejun Park, Elizabeth Harkavy, Guy Wuollet, Jeremy Zhang, Justin Thaler, Maggie Hsu, Miles Jennings, Pyrs Carvolth, Robert Hackett, Sam Broner, Scott Duke Kominers, Sean Neville, Shane Mac, and Sonal Chokshi
Translated by: Saoirse, Foresight News
This week, a16z released its annual "Big Ideas" report, featuring insights from partners across its Apps, American Dynamism, Bio, Crypto, Growth, Infra, and Speedrun teams. Below are 17 observations from multiple a16z crypto partners (including several guest contributors) on industry trends for 2026—covering agents and AI, stablecoins, tokenization and finance, privacy and security, extending into prediction markets, SNARKs and other applications, and concluding with reflections on industry building.
On Stablecoins, RWA Tokenization, Payments, and Finance
1. Better, More Flexible On-Ramps and Off-Ramps for Stablecoins
Last year, stablecoin transaction volume reached an estimated $46 trillion, continuously setting new historical highs. To put this in perspective, that’s over 20 times the transaction volume of PayPal and nearly three times that of Visa—one of the world’s largest payment networks—and rapidly approaching the volume of the U.S. Automated Clearing House (ACH), the electronic network handling direct deposits and other financial transactions in the United States.
Today, sending a stablecoin takes less than one second and costs under one cent in fees. Yet the unresolved core challenge remains: how to connect these “digital dollars” with the financial systems people use daily—the on-ramps and off-ramps for stablecoins.
A new generation of startups is filling this gap, integrating stablecoins with more widely adopted payment systems and local currencies: some leverage cryptographic proofs to let users privately convert local currency balances into digital dollars; others integrate regional networks using QR codes and real-time payment rails to enable bank-to-bank transfers; still others are building truly interoperable global wallet layers and card-issuing platforms that allow users to spend stablecoins directly at everyday merchants. Together, these solutions expand access to the digital dollar economy and may accelerate stablecoins’ adoption as mainstream payment tools.
As on- and off-ramps mature and digital dollars gain direct access to local payment systems and merchant tools, new use cases will emerge: cross-border workers can receive payments instantly, businesses can accept global dollars without needing a bank account, and apps can settle value with users worldwide in real time. At that point, stablecoins will transform from “niche financial instruments” into the “foundational settlement layer of the internet.”
—— Jeremy Zhang, a16z Crypto Engineering Team
2. Reimagining RWA Tokenization and Stablecoins Through a “Crypto-Native” Lens
Banks, fintech companies, and asset managers are showing strong interest in bringing traditional assets on-chain—including U.S. equities, commodities, indices, and other conventional assets. But as more traditional assets move on-chain, their tokenization often falls into the “digital mimicry trap”—merely replicating existing real-world asset structures instead of leveraging crypto-native advantages.
Synthetic derivatives like perpetual futures not only offer deeper liquidity but are also easier to implement. Additionally, perpetual contracts’ leveraged mechanics are intuitive, making them, in my view, the most product-market-fit crypto-native derivative. Emerging market equities are among the most suitable asset classes for “perpetualization” (for some stocks, zero-day expiry options already show higher liquidity than spot markets, making perpetual versions a valuable experiment).
This is fundamentally a choice between “fully on-chain vs. tokenized,” but regardless, in 2026 we’ll see more “crypto-native” approaches to RWA tokenization.
Similarly, stablecoins entered the mainstream in 2025, with outstanding issuance continuing to grow. In 2026, the stablecoin space will shift from “simple tokenization” to “innovative issuance models.” Today’s stablecoins lacking robust credit infrastructure resemble “narrow banks”—holding only highly secure, specific liquid assets. While narrow banking makes sense, it cannot serve as the long-term backbone of the on-chain economy.
Now, several new asset managers, fund managers, and protocols are exploring “on-chain lending backed by off-chain collateral,” though such loans are typically originated off-chain and then tokenized. I believe the value of tokenization here is minimal—it mainly serves users already in the on-chain ecosystem. Instead, debt assets should be “originated directly on-chain,” not “originated off-chain and later tokenized.” On-chain origination reduces servicing costs, back-end infrastructure expenses, and increases accessibility. Though compliance and standardization remain challenges, developers are actively working to solve them.
—— Guy Wuollet, General Partner, a16z Crypto
3. Stablecoins Drive Bank Ledger Upgrades, Unlocking New Payment Use Cases
Most banking software today is nearly unrecognizable to modern developers: banks were early adopters of large software systems in the 1960s–70s; second-generation core banking software emerged in the 1980s–90s (e.g., Temenos GLOBUS, Infosys Finacle). But these systems have aged, evolving extremely slowly—even today, the banking sector (especially core ledgers, the critical databases recording deposits, collateral, and other liabilities) often relies on mainframes running COBOL, using batch file interfaces instead of APIs.
The vast majority of global assets reside in these “decades-old core ledgers.” While these systems are battle-tested, regulatory-approved, and deeply embedded in complex banking operations, they severely hinder innovation: adding key features like real-time payments (RTP) can take months or even years, burdened by technical debt and regulatory complexity.
This is where stablecoins add value: over recent years, stablecoins achieved product-market fit and entered the mainstream. In 2025, traditional finance (TradFi) institutions fully embraced stablecoins. Stablecoins, tokenized deposits, tokenized Treasuries, and on-chain bonds now allow banks, fintechs, and financial institutions to build new products and serve new customers—crucially, without forcing them to overhaul legacy systems that, while outdated, have operated stably for decades. Stablecoins offer a low-risk path to innovation.
—— Sam Broner
4. The Internet Will Become the “Next-Generation Bank”
With the widespread adoption of AI agents, more commercial activity will happen “in the background” (rather than through user clicks), meaning the way value (money) moves must evolve accordingly.
In a world where systems act based on intent—not step-by-step instructions—such as AI agents automatically transferring funds upon identifying needs, fulfilling obligations, or triggering outcomes—value movement must match the speed and freedom of information flow today. Blockchain, smart contracts, and new protocols are key to achieving this.
Today, smart contracts can complete global dollar payments in seconds. By 2026, emerging foundational protocols like x402 will make settlement programmable and responsive: agents can instantly and permissionlessly pay for data, GPU compute, or API calls without invoicing, reconciliation, or batching; developer updates can embed payment rules, limits, and audit trails without fiat integration, merchant onboarding, or reliance on banks; prediction markets can “settle in real time” as events unfold—odds update, agent trades execute, global payouts finalize within seconds, without custodians or exchanges.
When value moves this way, “payments” cease to be a separate operational layer and become “network behavior”: banks will integrate into internet infrastructure, and assets will become infrastructure. If money flows like “internet-routable data packets,” the internet won’t just support financial systems—it will become the financial system itself.
—— Christian Crowley, Pyrs Carvolth, a16z Crypto Market Development Team
5. Wealth Management for Everyone
Traditionally, personalized wealth management has been reserved for banks’ “high-net-worth clients”: customized advice across asset classes and portfolio adjustments are costly and operationally complex. But as more asset classes become tokenized, crypto rails enable personalized strategies driven by “AI recommendations + assisted decision-making” to be executed instantly and rebalanced at low cost.
This goes beyond “robo-advisors”: everyone can access “active portfolio management” (not just passive). In 2025, traditional institutions increased crypto allocations in portfolios (banks advised 2%-5% directly or via exchange-traded products [ETPs]), but this is just the beginning. In 2026, we’ll see platforms focused on “wealth accumulation” (not just preservation) rise—fintechs like Revolut and Robinhood, and centralized exchanges like Coinbase, will leverage their tech stack advantages to capture this market.
Meanwhile, DeFi tools like Morpho Vaults can automatically allocate assets to lending markets offering optimal risk-adjusted returns, providing “core yield allocation” for portfolios. Holding idle liquidity in stablecoins (instead of fiat) or tokenized money market funds (instead of traditional ones) further expands yield potential.
Finally, tokenization enables retail investors to access “illiquid private market assets” (like private credit, pre-IPO equity, private equity) while meeting compliance and reporting requirements. Once balanced portfolios include tokenized assets—from bonds to equities to private and alternative assets—rebalancing can happen automatically, without wire transfers.
—— Maggie Hsu, a16z Crypto Market Development Team
On Agents and AI
6. From KYC to KYA
The bottleneck of the “agent economy” is shifting from intelligence to identity.
In financial services, non-human identities (like AI agents) already outnumber human employees by 96x—but these identities remain “ghosts unable to access banking systems.” The missing foundational capability is KYA (Know Your Agent).
Just as humans need credit scores to get loans, agents need “cryptographic attestation credentials” to transact—credentials linking the agent to its principal, constraints, and accountability. Without solving this, merchants will continue blocking agents at the firewall level. Industries that spent decades building KYC infrastructure now face solving KYA in months.
—— Sean Neville, Co-founder of Circle, Architect of USDC, CEO of Catena Labs
7. AI Empowers “Substantive Research Tasks”
As a mathematical economist, in January 2025 I struggled to get consumer-grade AI models to understand my workflow; by November, I could assign abstract tasks to AI models much like instructing a PhD student—sometimes receiving results that were innovative and correctly executed. Beyond my experience, AI adoption in research is growing, especially in reasoning: AI not only assists discovery directly but can autonomously solve Putnam problems—widely considered the hardest undergraduate math competition globally.
What remains to be explored is where such research assistance adds the most value and how to apply it. I expect AI will foster and reward a “new polymathic research paradigm”—emphasizing abilities like “speculating connections between ideas” and “rapidly deriving insight from highly speculative answers.” These answers may be inaccurate but point in the right direction (at least within certain logical frameworks). Ironically, this means “leveraging the power of model hallucinations”: when models are intelligent enough, giving them abstract exploration space may generate nonsense—but also breakthrough discoveries, much like humans being most creative in nonlinear, goal-ambiguous states.
Realizing this requires “new AI workflows”—not just agent-to-agent interaction, but “agents nesting within agents”: layered models help researchers evaluate prior models’ methods, progressively filtering useful from useless outputs. I’ve used this to write papers; others use it for patent searches, novel art creation, and (regrettably) discovering new smart contract attack vectors.
However, running “nested reasoning agent clusters” for research requires solving two key issues: “interoperability between models” and “identifying and fairly compensating each model’s contribution”—where cryptography can provide solutions.
—— Scott Duke Kominers, a16z Crypto Research Team, Professor at Harvard Business School
8. The “Invisible Tax” on Open Networks
The rise of AI agents imposes an “invisible tax” on open networks, undermining their economic foundation. This stems from the growing misalignment between the internet’s “context layer” and “execution layer”: AI agents extract data from ad-supported websites (context layer), providing user convenience while systematically bypassing revenue sources (ads, subscriptions) that sustain content creation.
To avoid open network decline (while protecting the diverse content fueling AI), we need large-scale “technical + economic” solutions—like next-gen sponsored content, micro-attribution systems, or new funding models. Current AI licensing deals are financially unsustainable stopgaps—compensating content providers with only a fraction of the revenue lost due to AI-driven traffic diversion.
Open networks need “new techno-economic models enabling automatic value flow.” The key 2026 shift: moving from “static licensing” to “real-time, usage-based payments.” This means testing and scaling blockchain-based micropayments + precise attribution standards—to automatically reward all parties contributing to an agent’s task completion.
—— Elizabeth Harkavy, a16z Crypto Investment Team
On Privacy and Security
9. Privacy as Crypto’s Most Important Moat
Privacy is essential for “global finance on-chain,” yet nearly all blockchains today lack it—on most chains, privacy is an afterthought.
Today, privacy capability alone can differentiate a chain from competitors. More importantly, privacy creates “chain lock-in effects,” or “privacy network effects”—especially crucial now that performance alone is no longer sufficient.
Thanks to cross-chain bridges, migrating between chains is easy when data is public. But with privacy, it’s different: “moving tokens across chains is easy; moving secrets across chains is hard.” Observers on chains, mempools, or network traffic can identify users when entering/exiting “private zones.” Transferring assets between a “privacy chain and a public chain,” or even between two privacy chains, leaks metadata like timing and amount correlations, increasing traceability risks.
Currently, many undifferentiated new chains compete down fees to near zero (on-chain space is effectively commoditized). Privacy-capable blockchains, however, can build stronger network effects. Reality is: unless a general-purpose chain has a thriving ecosystem, killer app, or unique distribution advantage, users and developers have no reason to choose it, build on it, or stay loyal.
On public chains, users easily trade across chains—choice of chain doesn’t matter. But on privacy chains, “which chain you choose” matters immensely: once users join a privacy chain, fear of exposing identity discourages migration, creating winner-take-all dynamics. Since privacy is essential in most real-world use cases, a few privacy chains may dominate crypto.
—— Ali Yahya, General Partner, a16z Crypto
10. The (Near) Future of Messaging: Quantum-Resistant and Decentralized
As the world prepares for the “quantum computing era,” encrypted messaging apps like Apple, Signal, and WhatsApp have taken action with notable success. But the issue is: all major messaging tools rely on “private servers operated by single entities”—vulnerable targets for governments to shut down, implant backdoors, or force disclosure of private data.
What’s the point of quantum-resistant encryption if a country can shut down servers, corporations hold private server keys, or corporations own the servers themselves? Private servers demand “trust me”; no private servers mean “you don’t need to trust me.” Messaging doesn’t need intermediaries (single corporations)—it needs “trustless open protocols.”
The path forward is “network decentralization”: no private servers, no single app, fully open-source code, using “top-tier cryptography” (including quantum resistance). In an open network, no individual, corporation, nonprofit, or nation can deny people communication rights—even if a country or company shuts down an app, 500 new versions appear overnight; even if a node is shut down, economic incentives from technologies like blockchain ensure new nodes immediately replace it.
Everything changes when people “control messages with keys” (like controlling funds): apps may evolve, but users retain control of messages and identity—even if they stop using an app, message ownership stays with them.
This isn’t just “quantum resistance” and “encryption”—it’s about “ownership” and “decentralization.” Without these, we’re building “unbreakable encryption that can be shut down at any moment.”
—— Shane Mac, Co-founder and CEO of XMTP Labs
11. “Secrets-as-a-Service”
Behind every model, agent, and automated system lies a simple foundation: data. Yet most data transmission channels—whether input to models or output from them—are opaque, tamperable, and unauditable. This may not matter for some consumer apps, but industries like finance and healthcare require enterprises to protect sensitive data privacy. It’s also a key barrier to institutional adoption of real-world asset tokenization.
How do we innovate securely, compliantly, autonomously, and with global interoperability while preserving privacy? Many solutions exist—I focus here on “data access control”: who controls sensitive data? How does data flow? Who (or what) has access rights?
Without data access control, any entity wanting to protect data confidentiality either relies on centralized services or builds custom systems—time-consuming, expensive, and limiting traditional institutions’ ability to fully leverage on-chain data management capabilities. As agent systems autonomously browse, transact, and decide, users and institutions across industries need “cryptographic-grade guarantees,” not “best-effort trust promises.”
That’s why I believe we need “Secrets-as-a-Service”: using new technologies to enable programmable native data access rules, client-side encryption, and decentralized key management—clearly defining who can decrypt which data, under what conditions, and for how long, with all rules enforced on-chain. Combined with verifiable data systems, “data confidentiality protection” becomes part of the internet’s foundational public infrastructure—not an afterthought patch at the application layer—making privacy a core infrastructure component.
—— Adeniyi Abiodun, Chief Product Officer and Co-founder, Mysten Labs
12. From “Code Is Law” to “Specifications Are Law”
Recent DeFi hacks have targeted long-proven protocols with strong teams, rigorous audits, and years of stable operation. These incidents reveal an unsettling truth: mainstream security practices remain largely reactive and case-by-case.
To mature DeFi security, two shifts are needed: from “patching vulnerability patterns” to “guaranteeing design-level properties,” and from “best-effort defense” to “principle-based systemic defense.” Two approaches:
Static/pre-deployment phase (testing, auditing, formal verification): Systematically prove “global invariants”—core rules the entire system always follows—rather than verifying only manually selected local rules. Multiple teams now develop AI-assisted proof tools that help write specifications, propose invariant hypotheses, and drastically reduce manual proof engineering work—which was previously too costly for broad adoption.
Dynamic/post-deployment phase (runtime monitoring, enforcement): Convert these “invariant rules” into real-time safety barriers, serving as the final line of defense. These barriers are coded as “runtime assertions”—transactions execute only if they satisfy assertion conditions.
Thus, we no longer assume “all vulnerabilities are fixed.” Instead, code itself enforces key security properties—any transaction violating them is automatically rejected.
This isn’t theoretical. Nearly all past hacks would have triggered such checks during execution, potentially preventing attacks. Thus, the once-popular “code is law” ideal is evolving into “specifications are law”: even under novel attacks, attackers must obey core security properties maintaining system integrity—leaving only trivial or highly difficult attack vectors.
—— Daejun Park, a16z Crypto Engineering Team
On Other Industries and Applications
13. Prediction Markets: Bigger, Broader, Smarter
Prediction markets have entered the mainstream. In 2026, deeper integration with crypto and AI will further scale, broaden, and enhance them—while posing new, significant challenges for developers to solve.
First, more contracts will launch. We’ll not only get real-time odds on “major elections, geopolitical events” but also on niche outcomes and complex cross-events. As these new contracts continuously release information and integrate into news ecosystems (a trend already visible), society faces key questions: how to balance this information’s value? How to improve transparency and audibility through better design (achievable via crypto)?
To handle surging contract volumes, new “consensus mechanisms” are needed for settlement. Centralized platforms (verifying event occurrence) matter, but controversial cases like “Zelenskyy lawsuit markets” or “Venezuela election markets” expose their limits. To resolve edge cases and expand into practical use, novel decentralized governance and LLM oracles can help verify disputed outcomes.
Beyond LLM oracles, AI unlocks further potential. AI agents trading on prediction platforms can gather diverse signals for short-term edges, offering new ways to understand the world and forecast trends (projects like Prophet Arena show promise). These agents can act as “advanced political analysts” whose self-developed strategies help us uncover core factors driving complex social events.
Will prediction markets replace polls? No. Instead, they’ll improve poll quality (poll data can feed into prediction markets). As a political scientist, I’m most excited about prediction markets co-evolving with a “rich, vibrant polling ecosystem”—enabled by new tech: AI can optimize survey experiences; crypto can offer new ways to prove poll respondents are real humans, not bots.
—— Andrew Hall, a16z Crypto Research Advisor, Professor of Political Economics at Stanford University
14. The Rise of Staked Media
Traditional media touts “objectivity,” but its flaws are well known. The internet gave everyone a voice—now more practitioners, builders, and doers speak directly to the public, reflecting their “skin in the game.” Ironically, audiences respect them not “despite their stakes” but “because of their stakes.”
The new shift isn’t social media’s rise, but the emergence of “crypto tools” enabling “publicly verifiable commitments.” As AI slashes content creation costs and enables effortless generation from any (real or fake) perspective, mere statements (human or bot) are no longer convincing. Tokenized assets, programmable locks, prediction markets, and on-chain histories offer stronger trust foundations: commentators can prove “walking the talk” (staking capital behind views); podcasters can lock tokens to prove they won’t flip-flop or “pump and dump”; analysts can tie predictions to “publicly settled markets,” creating auditable track records.
This is the early form of “staked media”: media embracing “skin in the game” with verifiable proof. Here, credibility comes not from “pretending neutrality” or “unsupported claims,” but from “publicly transparent, verifiable skin in the game.” Staked media won’t replace other forms but complement the ecosystem. It sends a new signal: not “believe me, I’m neutral,” but “this is the risk I’m willing to take, and here’s how you can verify my claims.”
—— Robert Hackett, a16z Crypto Editorial Team
15. Crypto Offers “New Foundational Primitives Beyond Blockchain”
For years, SNARKs—a cryptographic proof system allowing computation verification without re-execution—were mostly confined to blockchain use. The main reason: high cost. Generating a proof required up to a million times more work than executing the computation. Only in scenarios where costs could be amortized across thousands of validators (like blockchains) did it make sense—otherwise impractical.
But this is changing. By 2026, zkVM provers will cost around 10,000x (proof generation requiring 10,000x the work of direct computation), with memory usage in hundreds of megabytes—fast enough to run on phones, cheap enough for broad use. 10,000x is a key threshold because high-end GPUs offer about 10,000x the parallel processing power of laptop CPUs. By end-2026, a single GPU will “generate proofs for CPU computations in real time.”
This realizes visions from old research papers: “verifiable cloud computing.” If you run CPU workloads in the cloud due to insufficient compute, lack of expertise, or legacy system constraints, you’ll soon pay a modest premium to get cryptographic proof of correct computation. Provers are GPU-optimized—your code works without adaptation.
—— Justin Thaler, a16z Crypto Research Team, Associate Professor of Computer Science at Georgetown University
On Industry Building
16. Trading as a “Waystation,” Not a “Destination” for Crypto Companies
Today, aside from stablecoins and a few core infrastructure players, nearly every successful crypto company either runs or is transitioning into trading. But if “every crypto company becomes a trading platform,” where does that lead? Crowding into one lane fragments attention and leads to “few giants monopolize, most die.” This means companies rushing into trading miss chances to build “more competitive, sustainable business models.”
I understand founders’ drive for profitability, but “short-term product-market fit” has costs. This is acute in crypto: token dynamics tied to speculation tempt founders toward “instant gratification” paths during product-market fit—akin to the “marshmallow test” of delayed gratification.
Trading isn’t inherently bad—it’s a vital market function—but shouldn’t be the end goal. Founders focused on the “product essence” of product-market fit are ultimately more likely to win.
—— Arianna Simpson, General Partner, a16z Crypto
17. Unlocking Blockchain’s Full Potential: When Legal and Technical Architectures Align
Over the past decade, one of the biggest hurdles to building blockchain networks in the U.S. has been “legal uncertainty.” Broad and inconsistent enforcement of securities laws forced founders into regulatory frameworks designed for enterprises, not networks. For years, “avoiding legal risk” replaced “product strategy,” elevating lawyers over engineers.
This caused distortions: founders advised to avoid transparency; token distributions becoming legally arbitrary; governance turning ceremonial; organizations structured “to minimize legal risk first”; tokens deliberately “stripped of economic value” or “lacking business models.” Worse, projects “ignoring rules, operating in gray zones” often outpaced honest, compliant builders.
But now, the U.S. government is closer than ever to passing a “Crypto Market Structure Regulation Act”—which could eliminate all these distortions by 2026. If passed, it would incentivize transparency, establish clear standards, and replace “ad hoc enforcement” with “clear, structured paths for fundraising, token issuance, and decentralization.” After the GENIUS Act passed, stablecoin issuance surged; crypto market structure legislation will bring bigger change—focused squarely on “blockchain networks.”
In short, such regulation would let blockchain networks “operate truly as networks”: open, autonomous, composable, credibly neutral, and decentralized.
—— Miles Jennings, Member of a16z Crypto Policy Team, General Counsel
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
a16z
