Bitcoin mining difficulty fell 10.09% on June 14, 2026 — from 138.96 trillion to 124.93 trillion — marking the 11th largest single adjustment in Bitcoin's history and the second-largest drop of 2026. An estimated 100 exahashes per second (EH/s) went offline as publicly listed miners redirected po...
"Execution, not signing, becomes the next premium." — Matthew Sigel, Head of Digital Asset Research, VanEck
Bitcoin mining difficulty fell 10.09% on June 14, 2026 — from 138.96 trillion to 124.93 trillion — marking the 11th largest single adjustment in Bitcoin's history and the second-largest drop of 2026. An estimated 100 exahashes per second (EH/s) went offline as publicly listed miners redirected power, capital, and physical infrastructure toward artificial intelligence and high-performance computing workloads.
The numbers frame the shift. Public miners sold 32,000 BTC in Q1 2026 alone, a record quarterly liquidation, to fund data center conversions. Over $70 billion in cumulative AI and HPC contracts have been announced across the listed mining sector. Core Scientific's colocation revenue surged 800% year-over-year to $77.5 million in Q1 2026. IREN signed a five-year, $9.7 billion contract with Microsoft. Hut 8 entered a 15-year, $7 billion lease backed by Google. VanEck estimates the sector faces a $50 billion near-term funding gap to fulfill those commitments — and as much as $221 billion in long-term capital expenditure needs.
The weighted average cash cost to produce one bitcoin among listed miners hit approximately $79,995 in Q4 2025. Bitcoin traded between $62,000 and $70,000 through June 2026. The math forced the pivot. What remains unclear is whether the miners who signed AI contracts can actually build the infrastructure fast enough to collect on them.
At block height 953,568, Bitcoin completed its latest difficulty adjustment: a decrease of 10.09%, from 138.96T to 124.93T. The adjustment followed a period in which network hashrate dropped from approximately 1 zettahash per second (ZH/s) to around 900 EH/s, according to data from CoinWarz and Hashrate Index.
The drop confirms that a material portion of global hashrate stepped offline. Mining difficulty adjusts every 2,016 blocks (roughly two weeks) to maintain the 10-minute target block interval. A 10% decline means blocks were being found measurably slower than the protocol's target — consistent with roughly 100 EH/s of computing power going dark.
This was the second significant difficulty decline of 2026. The network had already seen a downward adjustment earlier in the year as miners began curtailing operations in response to post-halving margin compression. With the April 2024 halving cutting the block subsidy from 6.25 BTC to 3.125 BTC, revenue per terahash has been in structural decline.
Hashprice — the dollar revenue a miner earns per petahash per second per day — closed Q1 2026 at approximately $23.90/PH/s/day, according to CoinShares' Q1 2026 Bitcoin Mining Report. That figure represents the lowest reading since 2018.
The weighted average cash cost to produce one BTC among publicly listed miners reached approximately $79,995 in Q4 2025, according to CoinDesk reporting on quarterly filings. With Bitcoin trading in the $62,000–$70,000 band through Q2 2026, this implies losses of $10,000–$18,000 per coin mined for the average listed operator.
Profitability in June 2026 is confined to miners running current-generation hardware (sub-15 J/TH efficiency, such as the Antminer S21 class) with all-in electricity costs at or below $0.06/kWh. Industrial-scale operators meeting those thresholds report margins of 20–50%. Operators above $0.08/kWh are generally underwater. Home miners on residential power above $0.12/kWh are operating at a loss.
The network now consumes an estimated 350–420 GWh per day, or approximately 128 TWh annualized, with an average fleet efficiency of ~16 J/TH. According to industry data, 52.4% of mining electricity comes from zero-emission sources, up from 37.6% in 2022.
The scale of the transition is quantifiable. According to S&P Global research from February 2026 and subsequent industry disclosures, over $70 billion in cumulative AI and high-performance computing contracts have been announced across publicly listed Bitcoin miners.
As of Q4 2025, AI and HPC accounted for roughly 30% of revenue across listed miners on average. Analysts, including those at Bloomberg, project that figure could reach 70% by December 2026 for operators with executed contracts.
The economics driving the shift are straightforward. Bitcoin miners already control two inputs that AI infrastructure demands: large blocks of contracted power (measured in megawatts to gigawatts) and physical sites engineered for dense compute. When mining margins turn negative, diverting that capacity to AI workloads generating stable, contracted revenue becomes a financial imperative rather than a strategic choice.
The critical metric, according to VanEck's June 2026 framework report, is "gross energized power" — the actual megawatts a company has online and billing customers, as opposed to pipeline projections or signed-but-not-built capacity. Companies with energized AI capacity are being valued at 10x+ gross energized power. Those still relying on unbuilt pipeline trade at 2–6x.
Core Scientific (CORZ): Q1 2026 revenue of $115.2 million, with colocation revenue of $77.5 million — up from $8.6 million a year prior, an increase of approximately 800%. Colocation gross margin: 57%. The company has 243 MW of billed capacity, plans 450 MW by summer 2026, and targets 590 MW by early 2027. Core Scientific has secured over $10 billion in AI contracts with a 3 GW development pipeline, backed by a $3.3 billion financing round. It sold approximately $208 million worth of Bitcoin in Q1 to fund the transition. Net loss of $347.2 million in Q1 2026, driven primarily by $266.5 million in non-cash impairment charges on mining equipment.
IREN (formerly Iris Energy): Signed a five-year, $9.7 billion contract with Microsoft in November 2025 for 200 MW of liquid-cooled GPU infrastructure. Projected annualized run-rate revenue: $1.94 billion at 85% project-level EBITDA margin. IREN targets over $3.4 billion in annualized AI cloud revenue by end of 2026, supported by expansion to 140,000 NVIDIA GPUs across a 4.5 GW development pipeline.
Hut 8 (HUT): Entered a 15-year, $7 billion lease with Fluidstack (backed by Google) for initial deployment of 245 MW at its River Bend campus in Louisiana, with options extending to 2.295 GW. The deal includes commitments to serve Anthropic workloads. Google has taken an equity stake.
TeraWulf (WULF): Accumulated approximately $13 billion in contracted AI revenue across 10- to 25-year leases with Fluidstack and Core42. Google has backstopped $3.2 billion of those obligations and holds approximately 14% equity. VanEck classifies TeraWulf among the execution leaders based on energized capacity relative to contracted pipeline.
CleanSpark (CLSK): Pursuing a hybrid strategy — maintaining Bitcoin mining operations while building dedicated AI data centers. CleanSpark announced plans to construct facilities designed for both mining and AI workloads, attempting to maintain optionality across both revenue streams.
VanEck's June 16, 2026 research report, authored by analyst Griffin MacMaster and head of digital asset research Matthew Sigel, identified a collective near-term funding shortfall of approximately $50 billion across the listed mining sector. Long-term capital expenditure needs approach $221 billion.
The gap exists because the industry has delivered only about 25% of the AI and HPC capacity it has leased to customers. Contracts are signed; infrastructure is not built. Conversion from mining facilities to AI-grade data centers requires substantial capital for electrical upgrades, cooling systems (particularly liquid cooling for GPU clusters), network connectivity, and physical security compliance that cloud hyperscalers demand.
The report noted that investors are increasingly rewarding companies that have already energized AI capacity while penalizing those with large pipeline commitments but limited execution. Companies like Cipher Mining, Hut 8, and TeraWulf — with physical leases and energized power — command valuations above 10x gross energized power. Marathon Digital and CleanSpark, with limited contracted AI capacity, trade at 2–6x.
HIVE Digital Technologies faces the most acute funding strain relative to its market capitalization, driven by ambitions to deploy more than 100,000 GPUs in its "AI Gigafactory" program.
The capital requirements have triggered the largest wave of BTC selling by miners since the asset class's inception. According to Tekedia and CoinDesk reporting, public miners sold over 32,000 BTC in Q1 2026 — a record quarterly figure.
Specific disclosures include:
The trend represents a structural reversal of the "HODL" treasury strategy that defined the 2020–2024 era, when miners accumulated BTC on balance sheets as a leveraged bet on price appreciation. In 2026, BTC is being treated as working capital to be liquidated for operational needs — specifically, building AI infrastructure that generates contracted dollar revenue.
The migration of hashrate from Bitcoin mining to AI workloads has direct implications for network security. Hashrate and mining difficulty are the primary metrics of Bitcoin's resistance to 51% attacks. A 10% difficulty decline — while notable — does not approach a security threshold. The network remains at approximately 900 EH/s to 1 ZH/s, orders of magnitude above any theoretical attack vector.
However, the structural question is longer-term. If listed miners derive 70% of revenue from AI by year-end and continue curtailing mining operations, the network's hashrate growth trajectory changes. The 2024 halving already reduced the block subsidy. If transaction fees do not grow proportionally, the economic incentive to dedicate hardware to mining weakens relative to AI workloads.
The difficulty adjustment mechanism ensures the network continues functioning regardless of hashrate level. Blocks will still be found every ~10 minutes. But the concentration of remaining hashrate among fewer, larger operators — particularly those in low-cost jurisdictions — is a centralization vector worth monitoring.
The June 2026 difficulty adjustment is not an anomaly. It is a data point in a structural reallocation of capital and compute away from Bitcoin mining and toward AI infrastructure. The economics are clear: mining a bitcoin costs more than a bitcoin is worth for most listed operators, while AI colocation contracts offer contracted revenue at 57–85% gross margins.
The question is no longer whether Bitcoin miners will become AI companies. That transition is underway and quantified. The question is whether the $50 billion funding gap can be closed, whether the 25% delivery rate on contracted capacity will improve quickly enough to retain hyperscaler confidence, and whether the miners who signed the largest contracts can execute on infrastructure buildouts measured in gigawatts.
For Bitcoin's network, the implications are manageable in the short term but structurally significant. The incentive to mine is eroding for publicly listed operators. What remains of the hashrate will increasingly be provided by private operators, sovereign miners, and those with access to sub-$0.04/kWh power that makes mining viable regardless of AI alternatives. Bitcoin's security model is being stress-tested not by an attack, but by a better offer.