What Drove Bitcoin Below $70,000

Bitcoin's slide was the result of several converging headwinds that sapped institutional and retail confidence simultaneously.

• Sustained ETF outflows: Bitcoin spot ETFs had seen persistent negative flows since mid-May. Reports of a surprise sell-off from BlackRock's IBIT added significant downward pressure on price .
• The Saylor signal: Michael Saylor's Strategy (formerly MicroStrategy) sold $2.5 million worth of Bitcoin for the first time since 2022. This move was interpreted by the market as a potential crack in the thesis of the largest publicly traded BTC holder, spooking investors and pushing Bitcoin to its lowest level since April 7 .
• Whale distribution: On-chain data showed elevated large BTC transactions and net inflows to exchanges, suggesting that major holders were offloading positions .
• Liquidation cascade: The break of the $70,000 support triggered roughly $980 million in liquidations across the crypto market, compounding the sell-off in a classic leverage flush .

Why AI Tokens Bucked the Trend

While Bitcoin suffered, capital flowed decisively into tokens linked to artificial intelligence infrastructure. This rotation was not just speculative; it was backed by concrete catalysts.

Anthropic IPO and Nvidia Momentum

The filing for an initial public offering by AI company Anthropic, combined with new AI infrastructure announcements from Nvidia, created a powerful narrative tailwind. Tokens perceived as having a direct link to the AI buildout surged, with trading volumes exploding. NEAR Protocol, for instance, saw its 24-hour volume top $1 billion while Worldcoin and related assets saw comparable spikes .

A Classic Sector Rotation

Investor capital aggressively rotated out of Bitcoin ETFs and into AI-themed crypto assets. This was a flight toward a narrative that felt insulated from Bitcoin's immediate liquidity issues. Analysts attributed the move to capital concentrating in assets with clear AI narratives or tangible cash-flow expectations rather than pure monetary premiums . The appeal of a "secular trend" in AI, decoupled from short-term macro noise, made these tokens a relative safe haven for risk-on capital within a risk-off market .

The NEAR Short Squeeze

NEAR Protocol's rally was supercharged by specific technical factors. The token surged roughly 54% in a single week partly because short sellers were forced to cover their positions, absorbing millions of dollars in liquidations and accelerating the price increase. This short squeeze was amplified by a broader institutional interest narrative, including Bitwise's NEAR Staking ETP, and a product-specific story around chain abstraction .

A Structural Shift Toward Real-World Utility

The divergence between AI tokens and Bitcoin in early June is more than a temporary rotation—it marks a structural maturation of the crypto investor base.

• Valuing infrastructure over money: Investors are increasingly differentiating between Bitcoin's store-of-value thesis and tokens that power measurable infrastructure. Decentralized GPU computing networks like Render and io.net, AI-friendly Layer 1 blockchains like NEAR and ICP, and platforms for on-chain inference are being valued on developer activity, inference volume, and compute demand rather than macro correlation trades .
• Venture capital conviction: The capital flows are backed by long-term conviction. Venture capital allocation to AI-crypto hybrid startups reached 40% of all crypto venture funding in 2025, a massive jump from 18% the prior year, signaling that institutional money is betting on this convergence for the long haul .
• Decoupling as a feature: AI tokens are beginning to trade on their own fundamentals. This decoupling from Bitcoin's price action is a sign of a maturing asset class, where narratives with strong utility backing can attract liquidity independently .
• Underlying cost dynamics: The demand for decentralized compute is not theoretical. On-chain data and benchmarks suggest decentralized GPU networks can offer cost savings of 60-90% compared to traditional cloud providers like AWS and Azure for certain AI workloads, creating a genuine economic incentive for developers and enterprises .

A Critical Reality Check: Correlation and Bubble Risk

The structural shift is promising, but it does not eliminate risk. A 92% correlation remains between AI equities and AI crypto tokens, meaning a correction in the tech sector would likely trigger a sharp drawdown in these same crypto assets . The DeepSeek shock of January 2025, where a low-cost AI model dismantled the "expensive compute" thesis overnight, served as a brutal reminder that AI token valuations are sensitive to efficiency breakthroughs .

Investor concern is tangible. By early 2026, 54% of investors surveyed expressed fear of an AI bubble. However, analysts note a key difference from the dot-com era: much of the current buildout is funded by profitable tech giants' cash flows rather than speculative debt, which could provide a more durable floor .

The events of early June 2025 show a market learning to price utility, but also one still deeply intertwined with broader tech sentiment. For investors, the challenge is identifying the projects where on-chain activity and revenue justify the narrative before the next correlation event hits.

What Drove Bitcoin Below $70,000

Bitcoin's slide was the result of several converging headwinds that sapped institutional and retail confidence simultaneously.

• Sustained ETF outflows: Bitcoin spot ETFs had seen persistent negative flows since mid-May. Reports of a surprise sell-off from BlackRock's IBIT added significant downward pressure on price .
• The Saylor signal: Michael Saylor's Strategy (formerly MicroStrategy) sold $2.5 million worth of Bitcoin for the first time since 2022. This move was interpreted by the market as a potential crack in the thesis of the largest publicly traded BTC holder, spooking investors and pushing Bitcoin to its lowest level since April 7 .
• Whale distribution: On-chain data showed elevated large BTC transactions and net inflows to exchanges, suggesting that major holders were offloading positions .
• Liquidation cascade: The break of the $70,000 support triggered roughly $980 million in liquidations across the crypto market, compounding the sell-off in a classic leverage flush .

Why AI Tokens Bucked the Trend

While Bitcoin suffered, capital flowed decisively into tokens linked to artificial intelligence infrastructure. This rotation was not just speculative; it was backed by concrete catalysts.

Anthropic IPO and Nvidia Momentum

The filing for an initial public offering by AI company Anthropic, combined with new AI infrastructure announcements from Nvidia, created a powerful narrative tailwind. Tokens perceived as having a direct link to the AI buildout surged, with trading volumes exploding. NEAR Protocol, for instance, saw its 24-hour volume top $1 billion while Worldcoin and related assets saw comparable spikes .

A Classic Sector Rotation

Investor capital aggressively rotated out of Bitcoin ETFs and into AI-themed crypto assets. This was a flight toward a narrative that felt insulated from Bitcoin's immediate liquidity issues. Analysts attributed the move to capital concentrating in assets with clear AI narratives or tangible cash-flow expectations rather than pure monetary premiums . The appeal of a "secular trend" in AI, decoupled from short-term macro noise, made these tokens a relative safe haven for risk-on capital within a risk-off market .

The NEAR Short Squeeze

NEAR Protocol's rally was supercharged by specific technical factors. The token surged roughly 54% in a single week partly because short sellers were forced to cover their positions, absorbing millions of dollars in liquidations and accelerating the price increase. This short squeeze was amplified by a broader institutional interest narrative, including Bitwise's NEAR Staking ETP, and a product-specific story around chain abstraction .

A Structural Shift Toward Real-World Utility

The divergence between AI tokens and Bitcoin in early June is more than a temporary rotation—it marks a structural maturation of the crypto investor base.

• Valuing infrastructure over money: Investors are increasingly differentiating between Bitcoin's store-of-value thesis and tokens that power measurable infrastructure. Decentralized GPU computing networks like Render and io.net, AI-friendly Layer 1 blockchains like NEAR and ICP, and platforms for on-chain inference are being valued on developer activity, inference volume, and compute demand rather than macro correlation trades .
• Venture capital conviction: The capital flows are backed by long-term conviction. Venture capital allocation to AI-crypto hybrid startups reached 40% of all crypto venture funding in 2025, a massive jump from 18% the prior year, signaling that institutional money is betting on this convergence for the long haul .
• Decoupling as a feature: AI tokens are beginning to trade on their own fundamentals. This decoupling from Bitcoin's price action is a sign of a maturing asset class, where narratives with strong utility backing can attract liquidity independently .
• Underlying cost dynamics: The demand for decentralized compute is not theoretical. On-chain data and benchmarks suggest decentralized GPU networks can offer cost savings of 60-90% compared to traditional cloud providers like AWS and Azure for certain AI workloads, creating a genuine economic incentive for developers and enterprises .

A Critical Reality Check: Correlation and Bubble Risk

The structural shift is promising, but it does not eliminate risk. A 92% correlation remains between AI equities and AI crypto tokens, meaning a correction in the tech sector would likely trigger a sharp drawdown in these same crypto assets . The DeepSeek shock of January 2025, where a low-cost AI model dismantled the "expensive compute" thesis overnight, served as a brutal reminder that AI token valuations are sensitive to efficiency breakthroughs .

Investor concern is tangible. By early 2026, 54% of investors surveyed expressed fear of an AI bubble. However, analysts note a key difference from the dot-com era: much of the current buildout is funded by profitable tech giants' cash flows rather than speculative debt, which could provide a more durable floor .

The events of early June 2025 show a market learning to price utility, but also one still deeply intertwined with broader tech sentiment. For investors, the challenge is identifying the projects where on-chain activity and revenue justify the narrative before the next correlation event hits.