Publicly listed Bitcoin miners liquidated more than 32,000 BTC in Q1 2026 — more than in all of 2025 combined — to fund a mass migration into artificial intelligence infrastructure. The sector has signed over $70 billion in cumulative AI and high-performance computing (HPC) contracts, and CoinSha...
"We are no longer a Bitcoin company." — Ben Gagnon, CEO, Bitfarms
Publicly listed Bitcoin miners liquidated more than 32,000 BTC in Q1 2026 — more than in all of 2025 combined — to fund a mass migration into artificial intelligence infrastructure. The sector has signed over $70 billion in cumulative AI and high-performance computing (HPC) contracts, and CoinShares projects listed miners could derive 70% of their revenue from AI by year-end, up from approximately 30% in early 2026.
The pivot is not speculative. It is a rational response to collapsing mining economics: hashprice fell to an all-time post-halving low of $28–$30/PH/s/day in Q1 2026, the weighted average cost to produce one BTC among listed miners sits near $80,000, and 15–20% of the global fleet is estimated to be operating at a loss. Bitcoin's network hashrate dropped 23% from its October 2025 peak, triggering an 10.09% difficulty reduction in June — the second-largest downward adjustment of the year. The same megawatt allocated to mining generates volatile, shrinking revenue; leased to a hyperscaler under a 15-year AI contract, it delivers predictable, fixed-rate income at higher margins.
Miner equities have responded accordingly. TeraWulf (WULF) leads with a 73.58% YTD gain. Hut 8 (HUT) follows at 67.75%. Core Scientific (CORZ) is up more than 40%. All three are outperforming Bitcoin itself, which is down roughly 12% since January 1. The market is no longer pricing these firms on hash rate or BTC production. It is pricing contracted AI backlog, delivery timelines, and counterparty quality.
The April 2024 halving cut the block subsidy to 3.125 BTC, halving miners' guaranteed per-block revenue. Two years later, the math has become untenable for most operators.
According to CoinShares' Q1 2026 Bitcoin Mining Report, hashprice — the standard measure of mining revenue per unit of computing power — collapsed to approximately $28–$30/PH/s/day by early March, a new all-time post-halving low. The weighted average cash cost to produce one BTC among public miners reached approximately $80,000 in late 2025, and the figure has continued to rise as difficulty fluctuates.
Profitability now requires sub-$0.08/kWh power costs and hardware efficiency below 15 J/TH. Industrial-scale operators meeting both thresholds run margins of 20–50%. Home miners paying $0.12+/kWh are generally unprofitable. Hardware above 25 J/TH has been largely phased out by post-halving economics.
The competitive calculus is straightforward. CleanSpark President and CFO Gary Vecchiarelli stated during the company's Q1 2026 earnings call: "Right now [investing more in Bitcoin mining] doesn't make sense. So we want to be redirecting every dollar possible toward AI capex."
TeraWulf CEO Paul Prager framed the identity shift directly: "We are fundamentally a power company that builds digital infrastructure, not the other way around."
Bitcoin's network hashrate peaked above 1,000 EH/s earlier in 2026. It now stands at approximately 886 EH/s, a 23% decline from the October 2025 high and a 12% drop in June alone.
The drop triggered a 10.09% difficulty reduction at block 953,568 on June 14, taking difficulty from 138.96 trillion to 124.93 trillion. This was the second-largest downward adjustment of 2026 and the 11th-largest in Bitcoin's history.
Three factors drove the decline:
Price deterioration. Bitcoin retreated from approximately $81,000 in early May to $63,865, a 21% decline that compressed already thin margins below break-even for a significant portion of the fleet.
AI reallocation. Several public miners unplugged ASIC rigs or slowed mining expansion to retrofit sites for contracted AI/HPC workloads. The same 1 MW generates higher, more predictable returns when leased to a hyperscaler under a long-term AI contract compared with mining bitcoin at current hashprice levels.
Seasonal curtailment. Texas's four-coincident-peak (4CP) season began in June, creating incentives for large ERCOT-connected miners to curtail during peak intervals. This is a recurring seasonal pattern, but it compounded the structural decline.
The difficulty cut provides a temporary margin boost: surviving miners receive approximately 11% more BTC per unit of active hashrate. CoinShares still forecasts hashrate recovery to 1.8 ZH/s by year-end 2026, conditional on Bitcoin recovering toward $100,000. That forecast looks increasingly aspirational given current price levels.
Q1 2026 marked the first quarterly hashrate decline since Q1 2020, breaking five consecutive years of double-digit growth.
The scale of the AI pivot is measured not in megawatts but in contracted backlog. Over $70 billion in cumulative AI and HPC contracts have been signed across the public mining sector.
| Company | Contract Value | Counterparty | Duration | Capacity | |---------|---------------|--------------|----------|----------| | IREN (Iris Energy) | $3.4B managed cloud + $1.6B Dell GPU purchase | Dell / managed AI cloud | 5 years | 3 GW pipeline | | Core Scientific (CORZ) | $10.2B | CoreWeave | 12 years | 590 MW | | TeraWulf (WULF) | $12.8B | Core42, FluidStack (Google-backed) | Various | 510 MW critical IT load | | Hut 8 (HUT) | $7.0B | River Bend campus lease | 15 years | Not disclosed |
Core Scientific's transformation is the most complete. The company reported Q1 2026 revenue of $115.2 million, of which $77.5 million — 67% — came from AI data center colocation. Bitcoin mining revenue fell to $30.1 million. The company raised $3.3 billion in a bond sale in April to accelerate AI data center construction. Six data centers are under development, leased to CoreWeave for 12 years.
IREN signed a $1.6 billion purchase agreement with Dell on May 26 for Blackwell GPU systems that will service its AI cloud contract, with commissioning targeted for early 2027. The company holds a $14 billion market capitalization, making it the scale leader among listed miners, though it faces scrutiny over conversion timelines for its massive 3 GW power pipeline.
TeraWulf reported $21 million in HPC leasing revenue in Q1 2026, up more than 100% from Q4 2025, with contracted tenants including Core42 and Google-backed FluidStack.
The capital requirements for AI infrastructure have forced miners to liquidate their bitcoin treasuries at unprecedented rates.
According to CoinDesk, public miners collectively sold more than 32,000 BTC in Q1 2026 — a record that exceeds total BTC sales across the full year 2025. Core Scientific sold roughly 1,900 BTC worth $175 million in January alone and planned to liquidate substantially all remaining holdings during the quarter. Bitdeer reduced its BTC treasury to zero in February. Riot Platforms sold 1,818 BTC worth $162 million in December 2025.
The selling pattern reflects a fundamental change in how these companies view bitcoin. It is no longer a strategic reserve asset. It is working capital to be liquidated as needed to fund GPU purchases, data center construction, and grid interconnections.
Miners are also tapping debt markets. Core Scientific's $3.3 billion bond offering in April was structured to fund AI data center buildout. The shift from equity-funded BTC accumulation to debt-funded AI infrastructure represents a structural change in the sector's capital allocation.
Public markets have endorsed the pivot. According to Bitcoin.com News reporting, miner equities have outperformed BTC itself by approximately 70% in 2026:
| Company | Ticker | YTD Performance (2026) | |---------|--------|----------------------| | TeraWulf | WULF | +73.58% | | Hut 8 | HUT | +67.75% | | Core Scientific | CORZ | +40%+ | | Bitcoin (BTC) | — | –12% (approx.) |
Hut 8 trades at $77.06, the highest share price among the top ten listed miners by market valuation.
The re-rating is driven entirely by AI contract value. Miners without meaningful HPC contracts have underperformed. The market is pricing contracted backlog, delivery timelines, and counterparty quality — not hash rate or BTC production targets.
S&P Global's February 2026 research noted that the cryptocurrency market slump accelerated the pivot, but the structural incentive was already in place: long-term, fixed-rate AI infrastructure contracts offer fundamentally different risk-return profiles than volatile mining revenue.
The AI migration raises questions about Bitcoin's long-term security model.
Six mining pools currently control 99% of network hashrate. ASIC production is dominated by three firms — Bitmain, MicroBT, and Canaan. The combination of concentrated pool control and concentrated hardware supply creates supply chain fragility.
As large, listed miners redirect capacity to AI, their share of total hashpower diminishes. This could, paradoxically, improve geographic and operational diversity if smaller, more distributed operators fill the gap. But it could also reduce the total capital invested in Bitcoin security at a time when the block subsidy continues to decline.
The network's security budget remains dependent on transaction fees supplementing the reduced subsidy. If hashrate fails to recover toward the 1.8 ZH/s target CoinShares projects, Bitcoin's security model will face increasing scrutiny, particularly from institutional allocators performing network risk assessments.
The Bitcoin mining industry is undergoing its most significant structural transformation since the ASIC transition of 2013–2014. The difference is directional: that transition brought more capital into mining. This one is taking capital out.
The economics are unambiguous. At current BTC prices and hashprice levels, pure-play mining is a negative-margin business for a significant portion of operators. AI infrastructure offers structurally superior economics: long-term contracts, predictable revenue, higher margins. The market has ratified this assessment through stock prices.
The open question is what this means for Bitcoin's network. A 23% hashrate decline and record difficulty drops are manageable in the short term. The network's difficulty adjustment mechanism is designed precisely for this scenario. But the long-term trajectory — where the industry's largest, best-capitalized operators systematically reallocate capital away from mining — implies a security model that will rely increasingly on smaller operators, geographic diversity, and eventually, a robust transaction fee market.
For now, the listed miners are no longer mining companies. They are power and infrastructure companies that happen to still mine some bitcoin. The transition is not a pivot. It is a migration.