From Crypto Gold Rush to AI Mania: Why Venture Capital Left Web3, What It Still Funds, and What Comes Next

For a few brief years, crypto was the place where venture capital wanted to look early, brave, and slightly dangerous. Every major fund needed a Web3 thesis. Every generalist investor wanted access to token networks, DeFi infrastructure, NFT platforms, crypto gaming, Layer 1 chains, wallets, custodians, exchanges, and the mysterious possibility that the internet itself was being rebuilt with ownership baked in. Then the center of gravity shifted. After the collapse of Terra, Three Arrows Capital, FTX, BlockFi, Celsius, and the speculative excess around NFTs and token launches, venture capital did not disappear. It rotated. The new obsession became artificial intelligence.

The popular version of the story is simple: VCs used to fund crypto, then they abandoned it for AI. That story is directionally true, but incomplete. Venture capital did not fully leave crypto. It became more selective, more institutional, more infrastructure-focused, and less willing to underwrite abstract token narratives. At the same time, AI absorbed the oxygen in the room. The biggest checks, the best media narratives, the strongest founder migration, and the most aggressive limited partner enthusiasm moved toward AI. Crypto went from being the frontier trade to being a more mature, more scrutinized, and more cyclical sector.

The result is not a dead crypto VC market. It is a different one.

The Crypto Venture Boom Was Real

To understand the migration, it is worth remembering how large the crypto boom actually became. In 2021 and 2022, venture capital treated crypto as a generational platform shift. Funds were not merely investing in companies that happened to use blockchains. They were buying into an entire architecture of the future: decentralized finance, tokenized communities, metaverse economies, decentralized social networks, DAOs, NFT marketplaces, gaming economies, scaling networks, bridges, wallets, analytics platforms, staking infrastructure, and custody providers.

Crypto venture fundraising itself exploded. Galaxy’s institutional crypto venture data showed that global crypto and blockchain venture funds raised $20.55 billion in 2021 and then a record $37.7 billion in 2022 across 262 funds. That was extraordinary not only in absolute terms but as a share of global venture fundraising. Crypto funds represented roughly 12.62% of global venture fundraising in 2022, a peak that now looks like the high-water mark of the last cycle.

The deal environment was just as aggressive. At the height of the boom, crypto companies could raise on narratives that assumed mass adoption was near. Layer 1 protocols raised huge rounds. NFT marketplaces were valued like the future of culture. Play-to-earn gaming studios pitched tokenized economies as the next frontier of consumer internet. Infrastructure projects raised before product-market fit because investors believed the picks-and-shovels layer would capture value once the next billion users arrived.

The logic was not irrational. Crypto had created real markets. Bitcoin and Ethereum were liquid, global, and large. DeFi had demonstrated permissionless financial primitives. Stablecoins were becoming one of the first blockchain use cases with obvious global utility. NFTs had broken into mainstream culture, however briefly. Coinbase had gone public in 2021. Venture investors saw the possibility of a new technology platform with financial rails built in.

But the boom also had deep flaws. Too much capital chased too many similar ideas. Token incentives disguised weak retention. Some companies raised at valuations that assumed permanent bull-market liquidity. The distinction between community, speculation, and genuine demand became blurry. In hindsight, the crypto VC cycle of 2021–2022 was not just a funding boom. It was a narrative bubble.

The Crash Changed the Investor Psychology

The rotation away from crypto was not caused by one event. It was caused by a sequence of shocks that destroyed confidence at exactly the moment macro conditions became hostile to venture capital in general.

The Terra collapse damaged faith in algorithmic stablecoin design and wiped out enormous amounts of paper wealth. Three Arrows Capital showed how interconnected and leveraged crypto balance sheets had become. Celsius, BlockFi, Voyager, and other lenders exposed the fragility of centralized yield platforms that had marketed themselves as safe gateways into crypto returns. Then FTX collapsed, turning one of the industry’s most visible venture-backed companies into a symbol of governance failure.

This mattered enormously for venture capital because crypto investing depends on trust at several layers. Investors must trust the founders, the code, the custody setup, the token design, the market structure, and often the broader ecosystem in which the startup operates. After 2022, limited partners looked at crypto allocations with far more caution. A sector that had promised transparency had produced some of the most spectacular opacity in modern finance.

The data shows the damage. Crypto venture fundraising fell from $37.7 billion in 2022 to $5.75 billion in 2023 across 58 funds, according to Galaxy’s institutional crypto venture report. That was not a mild cooling. It was a collapse in allocator appetite. At the same time, global venture fundraising also contracted sharply, so crypto was hit by both sector-specific reputational damage and a broader venture downturn.

The macro backdrop made everything worse. Higher interest rates lowered the appeal of long-duration, speculative assets. Venture funds across sectors became more disciplined. Late-stage growth rounds slowed. IPO markets largely shut. Token markets recovered faster than private crypto valuations, but that did not automatically revive VC enthusiasm. Investors who had been burned by high-priced private rounds became reluctant to assume that rising Bitcoin prices would lift every Web3 startup.

That distinction is crucial. Bitcoin’s recovery did not mean crypto venture immediately recovered. Public tokens can reprice quickly. Private companies cannot. A VC-backed crypto startup that raised at a 2021 valuation still had to prove revenue, users, retention, and defensibility in a much tougher funding environment.

AI Became the New Center of Gravity

While crypto was digesting its failures, AI offered venture capital something it badly needed: a new platform story with immediate product pull.

The release of ChatGPT in late 2022 changed the market’s imagination. Suddenly, AI was not an abstract enterprise technology or a back-office automation tool. It was a consumer product, a developer platform, a productivity layer, and a possible operating system for knowledge work. Founders, engineers, enterprises, consumers, and investors all saw the shift at the same time.

For venture capital, AI had several advantages over crypto. It was easier to explain to limited partners. It had obvious enterprise demand. It could be deployed inside existing workflows. It did not require users to understand wallets, gas fees, bridges, custody, seed phrases, or tokenomics. It also had massive strategic buyers and partners: Microsoft, Google, Amazon, Meta, Nvidia, Salesforce, Oracle, ServiceNow, Adobe, Apple, and every cloud or enterprise software company with a balance sheet.

The funding numbers became enormous. CB Insights reported that AI represented 37% of global venture funding in 2024 and 17% of global venture deals, both all-time highs. Crunchbase data showed that AI captured close to 50% of all global venture funding in 2025, with roughly $202.3 billion invested in the sector, up more than 75% from about $114 billion in 2024. Foundation model companies alone raised around $80 billion in 2025, representing roughly 40% of global AI funding.

This is the kind of capital gravity crypto cannot currently match. In 2025, crypto and blockchain startups saw more than $20 billion in venture investment, according to Galaxy Research. That was a meaningful rebound and the strongest year since 2022. But AI was operating at a different scale. The comparison is not close: crypto recovered; AI dominated.

Did VCs Really Migrate From Crypto to AI?

The answer is yes, but not in the simplistic sense that venture capital packed up one office and moved to another.

Some crypto-native funds stayed crypto-native. They continued investing through the bear market, often in infrastructure, custody, compliance, stablecoins, scaling, security, and DeFi. Some even benefited from less crowded rounds. Meanwhile, many generalist funds that had aggressively entered crypto during the boom reduced their exposure, slowed deployment, or moved crypto from a core thesis to an opportunistic one.

The bigger migration happened at the margin. New founder attention moved toward AI. New analyst time moved toward AI. New LP curiosity moved toward AI. The prestige trade moved toward AI. The fastest fundraising stories moved toward AI. The largest late-stage rounds moved toward AI. A partner at a generalist fund who might have spent 2021 studying Layer 1 ecosystems or NFT marketplaces was, by 2024, more likely to be mapping foundation models, AI agents, infrastructure tooling, data pipelines, inference optimization, enterprise copilots, or vertical AI applications.

Galaxy itself noted in its 2025 crypto venture analysis that increased interest in artificial intelligence had diverted some attention from crypto investing. That sentence captures the phenomenon well. AI did not eliminate crypto venture. It competed for the same pool of risk capital, founder talent, media attention, and LP excitement.

The migration was also emotional. Crypto’s last boom ended in scandal and disappointment. AI’s boom began with product wonder. That contrast matters in venture capital, where conviction is often built around stories about the future. After 2022, crypto’s story became defensive: prove that the infrastructure matters, prove that the users are real, prove that the token is necessary, prove that the business can survive regulation. AI’s story was expansive: everything can be automated, every workflow can be rebuilt, every employee can be augmented, every software category can be reimagined.

Investors follow possibility. In 2023 and 2024, possibility wore an AI label.

The New Crypto VC Market Is Smaller, But More Serious

Although the boom-era euphoria is gone, crypto venture has not disappeared. In fact, the market has been rebuilding.

Galaxy Research reported that VCs invested $11.5 billion into crypto and blockchain startups across 2,153 deals in 2024. In 2025, that figure rose to more than $20 billion across 1,660 deals, the largest annual amount since 2022 and more than double 2023’s level. Q4 2025 alone saw $8.5 billion invested across 425 deals, the strongest quarterly capital deployment since Q2 2022.

But the composition of that funding changed. In 2024, early-stage deals still captured the majority of capital, while stablecoins, infrastructure, Web3, DeFi, and related categories attracted meaningful investment. By 2025, later-stage companies captured a record share of capital. Galaxy noted that 57% of crypto VC capital in 2025 went to later-stage companies, the largest share it had observed. That means investors are increasingly backing proven businesses rather than speculative early ecosystems.

The categories have also shifted. Trading, exchange, investing, and lending businesses attracted major capital in 2025, led by large rounds involving established firms such as Revolut and Kraken. Stablecoin companies became a major focus. Infrastructure remained important. Crypto AI appeared as a crossover theme, though it has been volatile and sometimes more narrative-driven than revenue-driven.

This is a much more mature pattern than the 2021 cycle. Instead of funding dozens of new Layer 1s, metaverse worlds, and NFT social platforms, VCs are asking harder questions: Where is the revenue? Where is the regulatory path? Where is the distribution? Does the token create real economic coordination, or is it just a fundraising device? Is this product better because it is on-chain, or merely more complicated?

That scrutiny is healthy. It also makes fundraising harder for founders who are selling ideology rather than traction.

Why Bitcoin’s Bull Market Did Not Fully Reignite Crypto VC

One of the most interesting features of the current cycle is the weakened relationship between liquid crypto prices and private crypto venture funding.

In earlier cycles, rising Bitcoin and Ethereum prices tended to generate more venture enthusiasm. Token wealth created new angel investors. Crypto funds raised more easily. Founders launched more projects. Retail attention spilled into startup narratives. The public market and private market moved together.

That relationship has been weaker since 2023. Bitcoin recovered sharply, spot Bitcoin ETFs brought major institutional legitimacy, and public crypto market capitalization improved. Yet private crypto venture activity lagged for much of the period. Galaxy’s 2024 research noted that the multi-year correlation between Bitcoin price and crypto VC investment had struggled to recover. Its 2025 research again described a weaker relationship, even though some correlation remained.

There are several reasons.

First, the current bull market has been highly Bitcoin-centric. Spot ETFs channeled institutional demand into Bitcoin itself, not necessarily into private crypto startups. A pension fund or wealth manager that wants crypto exposure can now buy a regulated ETF rather than commit to a ten-year illiquid venture fund.

Second, crypto treasury companies and public-market vehicles competed for institutional capital. Investors who wanted beta to the asset class had more liquid options than early-stage venture.

Third, many of the hottest 2021 categories did not recover in the same way. NFTs, metaverse land, play-to-earn gaming, and many DAO experiments lost credibility. That removed a large part of the venture imagination from the last cycle.

Fourth, the exit environment remains challenging. Crypto has produced public companies and acquisitions, but not enough predictable exits to satisfy all LPs. Venture capital needs distributions. A token listing can create liquidity, but regulatory uncertainty and changing market norms have made token-based exits less straightforward than during the boom.

Finally, AI became the better LP conversation. Even when crypto prices were strong, AI was where the largest private companies, biggest rounds, and clearest enterprise adoption stories were happening.

What VCs Fund in Crypto Now

Today’s crypto VC market is not funding “Web3” as a broad slogan. It is funding narrower, more defensible themes.

Stablecoins are one of the strongest categories. They are arguably crypto’s clearest product-market fit after Bitcoin. Dollar-backed tokens move value globally, operate around the clock, and are increasingly integrated into payments, remittances, trading, treasury management, and emerging-market finance. Venture investors like stablecoins because they combine crypto rails with familiar financial demand. They also have clearer revenue models than many tokenized consumer apps.

Infrastructure remains a core theme. This includes custody, wallets, security, data indexing, compliance tooling, developer platforms, institutional trading infrastructure, staking services, restaking systems, interoperability, and scaling technology. These businesses are less glamorous than consumer crypto, but they often serve real customers and can generate revenue regardless of retail hype.

Tokenization is another major area. Funds, treasuries, private credit, real estate, and other real-world assets are increasingly being explored on-chain. The thesis is that blockchains can improve settlement, transparency, transferability, and composability for financial assets. This area appeals to institutions because it looks less like speculative crypto culture and more like financial infrastructure modernization.

DeFi is still funded, but with more skepticism. Investors want protocols that solve specific liquidity, risk, trading, lending, or settlement problems. The era of funding yet another yield farm is over. The more interesting DeFi opportunities are in intent-based execution, derivatives, structured products, lending against real assets, stablecoin liquidity, and institutional-grade market infrastructure.

Consumer crypto remains difficult. Social, gaming, NFTs, creator economies, and decentralized identity still attract founders, but many VCs are cautious. The lesson from the last cycle is that token incentives can create activity without durable demand. A consumer crypto startup now needs to prove retention, not just transaction volume.

Crypto AI sits at the intersection of the two biggest narratives, which makes it attractive but dangerous. Some projects are genuinely useful: decentralized compute markets, AI agent payments, data provenance, model verification, autonomous wallets, and machine-to-machine settlement. Others simply attach an AI story to a token. VCs are interested, but increasingly aware that “AI plus crypto” is not automatically a business.

Privacy is also re-emerging as a category. As AI improves surveillance and blockchain analytics becomes more powerful, private transactions, confidential computing, and selective disclosure may become more important. This remains a regulatory-sensitive area, but the strategic need is real.

Why AI Attracts the Mega-Rounds

AI’s funding dominance is not just about hype. The capital requirements are structurally different.

Foundation model companies need extraordinary amounts of money for compute, talent, data, and infrastructure. Training and serving frontier models is expensive. This creates a market where companies can raise billions and still plausibly argue they need more. OpenAI, Anthropic, xAI, Mistral, Cohere, and other model companies are not raising like normal software startups. They are raising more like strategic infrastructure projects.

That changes the venture landscape. A few AI companies can absorb huge amounts of global capital. Crunchbase reported that OpenAI and Anthropic alone captured a significant share of global venture investment in 2025. This concentration makes AI’s funding numbers look massive, but it also means much of the money is flowing into a small number of companies.

Crypto does not currently have an equivalent. A blockchain infrastructure startup may need significant capital, but most do not require tens of billions in training compute. The biggest crypto rounds tend to involve exchanges, trading platforms, stablecoin issuers, or later-stage financial businesses rather than frontier research labs.

AI also benefits from strategic corporate participation. Cloud providers, chip companies, enterprise software giants, and large technology platforms have direct reasons to fund AI companies. They want cloud workloads, model access, strategic optionality, talent, and defensive positioning. Crypto has strategic investors too, but the corporate demand is narrower and often more regulated.

This is why AI can dominate venture even if many AI startups later fail. The largest players are seen as critical infrastructure for the next technology platform. That is the kind of story that unlocks huge checks.

Is AI Becoming the New Bubble?

The uncomfortable answer is probably yes, at least in parts of the market.

The parallels to crypto are obvious. There is a platform-shift narrative. There is intense founder migration. There are inflated valuations. There are companies raising before durable business models are proven. There is a wave of startups using the label to attract capital. There is fear of missing out among investors who do not want to explain to LPs why they missed the defining technology cycle of the decade.

But there is also an important difference. AI has already shown broad utility across coding, writing, customer support, design, research, sales, data analysis, cybersecurity, drug discovery, robotics, and enterprise workflow automation. Crypto’s strongest use cases are real, but narrower: store of value, stablecoin transfer, decentralized exchange, custody, settlement, token issuance, and financial experimentation. AI feels immediately useful to a much larger number of businesses.

That does not mean AI valuations are justified. Many AI application startups may be crushed by model providers, incumbents, open-source alternatives, or declining margins. Some model companies may never produce returns proportional to their capital consumption. The infrastructure buildout could overshoot. The same investors who overpaid for crypto networks in 2021 may now be overpaying for AI copilots in 2025.

The more precise point is that venture capital does not rotate away from bubbles because it dislikes them. Venture capital often needs bubbles. Bubbles create talent formation, infrastructure buildout, public attention, and risk appetite. The question is not whether AI has speculative excess. It does. The question is whether enough enduring companies will emerge from that excess. Investors believe the answer is yes.

What Crypto Can Learn From the AI Rotation

Crypto should not respond to the AI boom by trying to imitate its language. Slapping “AI” onto token projects will not fix the sector’s credibility problem. The better lesson is that users and investors reward obvious utility.

AI adoption exploded because people could try the product and immediately understand the value. Crypto still often asks users to endure complexity in exchange for ideological or financial upside. That is not enough for mainstream adoption.

The next generation of crypto startups must hide complexity. Wallets need to feel like normal accounts. Stablecoin payments need to feel cheaper and faster than bank transfers. Tokenization needs to make settlement better for institutions. DeFi needs to offer liquidity and transparency without exposing users to avoidable risk. On-chain identity and privacy need to solve real problems without becoming compliance nightmares.

VCs are still willing to fund crypto, but they are less willing to fund abstractions. The most attractive crypto companies now look more like infrastructure businesses, financial networks, compliance-aware platforms, payment rails, or developer tools. The least attractive ones look like recycled token narratives from the last cycle.

What Happens Next

The likely future is not a simple return to 2021. Crypto VC will recover, but it will not look like the last boom.

Funding should continue flowing into stablecoins, institutional infrastructure, tokenization, custody, compliance, security, scaling, and wallet UX. The strongest early-stage founders will still raise, especially if they can show why blockchains are necessary rather than decorative. Later-stage funding may remain concentrated in companies with revenue, regulatory positioning, and strategic importance.

Generalist funds will come back more aggressively if exits improve. That means more IPOs, acquisitions, token liquidity events with clearer compliance, and visible distributions to LPs. Venture is ultimately a returns business. If crypto can produce liquid winners again, capital will follow.

AI will remain the dominant VC theme for the near future. It has too much momentum, too much enterprise demand, and too much strategic capital behind it. But AI’s dominance may eventually help crypto rather than simply crowd it out. AI agents may need wallets. Autonomous software may need payment rails. Machine-to-machine transactions may use stablecoins. Provenance and verification may require blockchains. Decentralized compute and data markets may become more relevant if centralized AI infrastructure becomes too expensive or too concentrated.

The most interesting long-term opportunity may be the convergence of AI and crypto, but only where the combination is functional. AI does not need a token for every workflow. Crypto does not need AI for every protocol. But agents that transact, negotiate, pay, verify, and coordinate across digital systems may need open financial rails. That is where crypto can become infrastructure for the AI economy rather than a rival narrative.

Verdict: The Migration Was Real, But Crypto Is Not Over

The claim that venture capital moved from crypto to AI is broadly true. The data supports it. Crypto venture fundraising collapsed after the 2022 peak. AI captured a record share of global venture funding in 2024 and nearly half of global funding in 2025. Founder attention, LP excitement, mega-rounds, and generalist VC energy clearly shifted toward AI.

But the stronger interpretation is that crypto matured under pressure while AI entered its expansionary boom. Crypto is no longer the default frontier for risk capital. It has to compete on fundamentals. That is painful for weaker projects and healthy for the sector.

In the last cycle, crypto raised like a revolution. Today, it raises more like financial infrastructure. That may sound less exciting, but it is probably more sustainable.

VCs have not stopped investing in crypto. They have stopped funding every crypto story equally. The new money wants stablecoins, infrastructure, tokenization, custody, security, institutional access, real users, and revenue. It wants fewer white papers and more distribution. It wants fewer token games and more durable networks.

AI is now the brightest object in venture capital. Crypto used to be. One day, AI will also face its own correction, and investors will separate the enduring companies from the narrative passengers. Crypto has already gone through that humiliation. The sector that remains is smaller in ego, stronger in infrastructure, and more focused on use cases that actually matter.

The next crypto boom may not be driven by JPEGs, metaverse land, or another Layer 1 arms race. It may be driven by stablecoin settlement, tokenized assets, programmable finance, AI-agent payments, privacy, and institutional rails. If that happens, venture capital will not need to rediscover crypto as a fantasy of the future. It will fund it as part of the financial and computational infrastructure of the present.