MARA Holdings’ stock cooled after Tuesday’s session as the Bitcoin miner reported a sharply extended first-quarter loss and revenue that fell short of expectations. The results highlight the sector’s ongoing pressure from Bitcoin’s price moves and a challenging mining environment, even as MARA leans into a broader AI-focused growth strategy.

For the quarter ended March 31, MARA said revenue declined 18% year-over-year to $174.6 million, missing Wall Street estimates of about $192.7 million. The company swamped investors with a substantial net loss of $1.3 billion, compared with a $533.4 million loss in the prior-year quarter. Earnings per share came in at a negative $3.31, versus consensus expectations around a $2.20 per-share loss.

In after-hours trading, MARA shares slid about 3.4% to $12.93, erasing gains from the regular session, which finished up roughly 3.5% at $13.39. The stock has underperformed the broader year so far, with a roughly 16% drop over the past 12 months.

The quarterly loss was largely driven by unrealized losses on MARA’s Bitcoin treasury—38,689 BTC—amid a roughly 23% slide in the cryptocurrency over the period. MARA also disclosed that it sold more than 15,100 BTC worth about $1.1 billion in the final week of March, a move described as aimed at deleveraging by acquiring debt at a discount.

Despite the near-term pain, MARA reiterated its long-term strategy of anchoring operations in Bitcoin mining while expanding into artificial intelligence and high-performance computing (HPC) as new revenue streams. The company characterized Bitcoin mining as its “operational foundation,” even as it pursues AI-driven data center opportunities on the same sites or adjacent facilities.

Market conditions for Bitcoin mining remain tough. Bitcoin traded more than 35% below its all-time peak of $126,080, undermining miner revenue per block. At the same time, mining difficulty has risen by about 30% over the past year, heightening the hurdle for new and existing operations. Against this backdrop, MARA has slipped from being the largest Bitcoin miner by market capitalization to roughly seventh place as rivals push more aggressively into AI-related infrastructure.

MARA’s current AI strategy centers on a partnership with Starwood Capital to convert some Bitcoin mining sites into AI and HPC data centers, and the acquisition of Long Ridge Energy & Power—a gas-fired power plant and data center facility—for $1.5 billion in late April. The combination of these moves could reshape how the company monetizes its energy footprint over time. In its statements, MARA described a flexible operating model: it can continue generating revenue today from Bitcoin mining while preserving the option to redirect power toward AI and other IT workloads as opportunities mature on the same sites.

According to the company, the Long Ridge acquisition could ultimately support up to 600 megawatts of AI computing capacity, and about 90% of MARA’s non-hosted mining capacity could be redeployed for AI and IT compute. Notably, MARA also signaled it does not plan to purchase additional Bitcoin mining hardware in the near term, signaling that the near-term focus is on redeployable infrastructure and the AI/HPC push rather than expanding traditional mining capacity.

The evolving strategy comes as the broader market contends with a blend of macro headwinds and sector-specific pressures. The industry has been navigating tighter margins as price volatility and rising energy costs compress profits, even as some players look to diversify into data-center-enabled services. MARA’s pivot toward AI and HPC aligns with a wider trend among miners to monetize energy assets through adjacent digital infrastructure use cases when Bitcoin mining alone becomes less favorable.

Analysts who track the sector note that the transition from pure mining to AI-enabled data centers introduces new variables. Revenue visibility may improve if AI demand strengthens, but it also hinges on energy pricing, site uptime, and the pace of customer adoption for AI workloads. MARA’s disclosures suggest a careful, staged approach: keep Bitcoin mining running to generate cash flow today, while gradually repurposing sites for AI capacity as market conditions and technology maturity permit.

As the year unfolds, investors will be watching how effectively MARA can translate its physical assets into AI-ready capacity and how the company manages debt and liquidity in a capital-intensive deployment. With the LED of AI-driven compute on its sites, MARA faces a delicate balancing act between sustaining mining revenue and realizing the strategic upside from its data-center ambitions.

Readers should monitor the company’s quarterly updates for progress on Starwood Capital collaborations and the Long Ridge project’s development cadence, as well as any commentary on energy-price trends and Bitcoin’s price trajectory, which continue to be decisive for mining economics and the viability of the company’s dual-track strategy.

What’s next remains uncertain, but MARA’s emphasis on flexible infrastructure and multi-use sites could redefine how Bitcoin miners steward capital and resources if AI demand materializes alongside Bitcoin mining profitability.

SharpLink reported $12.1 million in first-quarter 2026 revenue, up from $742,000 in the same period last year. 

The company said the increase came mainly from its actively managed Ethereum treasury strategy, which began in June 2025.

The Nasdaq-listed firm now presents itself as an institutional Ethereum treasury platform. Its website says SharpLink is listed under the ticker SBET and gives investors a public-market vehicle tied to ETH exposure and yield.

ETH losses weigh on earnings

SharpLink still posted a net loss of $685.6 million in Q1, compared with a $1.0 million loss a year earlier. The company tied the loss mainly to non-cash ETH market losses and impairment charges during a weaker quarter for Ethereum.

The company reported $506.7 million in unrealized ETH losses and a $191.7 million LsETH impairment charge. It also said those accounting losses did not reduce the number of ETH held by the company.

Notably, SharpLink held about 870,821 ETH at the end of March. That figure rose to 872,984 ETH as of May 4, keeping the company among the largest public Ethereum treasury holders.

The company has also generated 18,800 ETH in staking rewards since June 2025 through native and liquid staking programs. Earlier comments to crypto.news showed that SharpLink had already planned to move beyond basic staking into restaking, lending, and other Ethereum-based yield tools.

Chief executive Joseph Chalom said, “We’re trying to hit singles and doubles.” He also said the company is not seeking VC-like returns, framing the plan as a lower-risk yield strategy rather than a hunt for aggressive gains.

Galaxy fund brings next test

SharpLink and Galaxy Digital plan to launch the Galaxy SharpLink Onchain Yield Fund with about $125 million in commitments. SharpLink said the fund will deploy capital into selected onchain opportunities and provide liquidity to emerging protocols.

Chalom said the Galaxy partnership will use institutional-grade strategies to provide liquidity to high-quality protocols while seeking returns for shareholders. The company also warned that the fund may not launch on schedule, commitments may not be funded, and strategies may produce losses.

Related market updates show that more public companies are building ETH treasury models. FG Nexus, for example, disclosed 47,331 ETH in holdings and said it planned to use staking, restaking, and DeFi markets to earn yield.

TLDR

• Financial educator Robert Kiyosaki predicts a devastating economic collapse in 2026 driven by unsustainable debt levels
• Silver ranks as his number one investment choice, a position he’s held since purchasing the metal in 1965
• With silver currently hovering around $85 per ounce, Kiyosaki projects a rise to $200
• His complete portfolio for weathering the crisis includes gold, silver, energy, food production, Bitcoin, and Ethereum
• Market experts corroborate his silver thesis, pointing to depleted exchange inventories and surging industrial consumption

Robert Kiyosaki, renowned for his bestselling book Rich Dad Poor Dad, has issued a dire forecast: the world economy faces a catastrophic downturn in 2026. According to Kiyosaki, this collapse will devastate those caught unprepared while enriching investors positioned in tangible assets.

The financial educator attributes the looming crisis to America’s staggering $39 trillion national debt combined with ongoing currency devaluation that he claims began in 1974. He also identifies vulnerable retirement portfolios held by the baby boomer generation as a critical weak point.

Kiyosaki refers to this phenomenon as the “Everything Bubble,” a concept he introduced in his 2002 publication Rich Dad’s Prophecy. According to his analysis, that bubble has reached its breaking point.

“In 2026 the global economy is about to crash. That’s good news for those that can see the future. Bad news for the blind,” Kiyosaki declared on X.

Conventional financial organizations largely disagree with this assessment. Most international economic forecasters continue to anticipate steady growth through 2026, though they acknowledge elevated risks related to government debt and international tensions.

Kiyosaki maintains that previous market downturns in 1987, 2000, 2008, and 2022 actually increased his wealth because he maintained positions in physical assets. He intends to deploy identical tactics for the anticipated 2026 crisis.

His primary focus currently centers on silver. His investment journey with the precious metal began in 1965 when he was just 18 years old, purchasing it for mere pennies per ounce. Today, he characterizes it as among his most successful investment decisions.

Why Silver Stands Out to Kiyosaki

Silver spot prices are currently fluctuating near $85 per ounce, representing substantial gains over the previous twelve months. Kiyosaki maintains a long-range price projection of $200 per ounce.

He values silver for its dual nature as both a monetary protection and an industrial commodity. The metal plays essential roles in solar energy systems, electric vehicle production, battery technology, and artificial intelligence hardware.

The global silver market has experienced six consecutive years of supply shortfalls. Industrial applications now account for approximately 50% of worldwide silver consumption.

Additional market analysts echo his perspective. Trading veteran Vijay identified silver in the $75 to $80 range as exceptionally undervalued, highlighting CME warehouse levels at their lowest since January 2025.

Analytical firm World of Finance and Associates established a short-term resistance zone between $88 and $92 per ounce, absent significant economic disruptions. Several precious metals specialists have additionally highlighted silver mining companies as a magnified approach to capitalize on advancing prices.

Bitcoin Also on Kiyosaki’s Radar

Kiyosaki’s investment strategy for 2026 extends beyond silver. His portfolio also encompasses gold, energy resources, agricultural production, Bitcoin, and Ethereum as reliable holdings during monetary system deterioration.

He has revealed accumulating Bitcoin around the $67,000 level and previously established a 2026 price objective of $250,000 per coin. He positions Bitcoin and silver as synergistic protections against currency debasement.

His six-decade history with silver investments provides the foundation for his investment thesis. The S&P 500 has delivered approximately 400x returns over the comparable timeframe with dividend reinvestment, contrasted with silver’s roughly 63x appreciation. Skeptics reference this performance differential when challenging his approach.

Nevertheless, Kiyosaki demonstrates no indication of altering his strategy. He concluded his latest statement with a pointed question to his audience: “What do you see happening in the future? What can you invest in?”

Bitcoin’s BTC$81,002.49 transparency was once considered one of its greatest strengths. Now, Ray Dalio says, it may be the very reason central banks won’t adopt it as a reserve asset, even though corporations and institutional investors have embraced it.

The billionaire hedge fund manager, who is also a bitcoin investor, said on X that, “Bitcoin lacks privacy. Transactions can be monitored and potentially controlled, which is why central banks aren’t looking to hold it.”

Ray Dalio has previously said he allocates about 1% of his portfolio to bitcoin.

Bitcoin, the world’s largest blockchain network, operates as a decentralized peer-to-peer system built on a public ledger. Every transaction is permanently recorded on this transparent ledger, allowing anyone to view it in real time.

Anyone can open a Bitcoin block explorer, enter a wallet address into the search bar, and view the entire transaction history associated with it. While wallet addresses are pseudonymous rather than directly tied to identities, blockchain analytics firms and law enforcement agencies can often trace the movement of funds and link activity back to individuals or institutions.

In other words, the flow of BTC, the blockchain’s native token, is highly transparent and traceable, even if it is not always directly tied to real-world identities.

This level of transparency, often praised by Bitcoin supporters, may also be what keeps central banks away. Imagine being a central bank and accumulating an asset whose flows can be tracked in real time on a public ledger.

The lack of privacy is also a concern for large institutional players. At Consensus Hong Kong in February, participants noted that the mass adoption of blockchain technology at the institutional level may ultimately depend on stronger privacy features, particularly for large transactions.

The market seems to align with the growing expert consensus on privacy. For instance, the privacy-focused coin zcash (ZEC) has surged over 800% since early 2025. Bitcoin, meanwhile, is down over 10%.

Correlated to stocks

Dalio’s concerns, however, go beyond central bank adoption. He pointed to structural issues that limit bitcoin’s appeal as a reserve asset compared to traditional alternatives like gold.

One of them is its tendency to take cues from Wall Street, especially the technology stocks, rather than acting as an independent store of value during periods of stress.

As of writing, the 90-day correlation coefficient between bitcoin and the Nasdaq, Wall Street’s tech-heavy index, was 0.89, according to data source TradingView. That translates into an R² of 0.79, meaning roughly 79% of bitcoin’s price movements can be explained by its relationship with the Nasdaq over the 90 days. The data points to BTC’s behavior more as a risk-on asset than an independent store of value.

The other issue Dalio highlighted is the market’s scale and structure. Unlike gold, which is deeply established, widely held, and exists outside any single digital system, bitcoin remains a relatively small and more easily influenced market. In his view, these factors further weaken its case as a global reserve asset, despite growing institutional participation.

“Ultimately, gold is more widely held, deeply established, and still plays a central role in the global system,” he said.

Dalio has repeatedly favored gold over bitcoin, and his views have been countered by crypto industry experts.

Google’s Threat Intelligence Group said it found a zero-day exploit that likely used artificial intelligence during discovery and weaponization. 

The exploit targeted a popular open-source, web-based system administration tool and allowed attackers to bypass two-factor authentication after gaining valid login details.

The group said it worked with the affected vendor to disclose the flaw and stop the planned mass exploitation campaign. Google did not name the tool, the vendor, or the threat actor behind the operation.

Exploit needed valid credentials first

The flaw did not give attackers full access on its own. Google said the bypass required valid user credentials before the attacker could skip the second login step. That detail matters because two-factor authentication often protects crypto accounts, exchange logins, developer dashboards, and wallet-linked services.

Google said the weakness came from a logic error, not a common coding bug such as memory corruption or poor input handling. The company described it as a high-level semantic flaw, where a hardcoded trust assumption conflicted with the tool’s 2FA checks.

Moreover, Google said it had “high confidence” that the actor likely used an AI model to support discovery and weaponization of the vulnerability. The company said the exploit script included educational comments, a hallucinated CVSS score, and a clean Python format often linked to large language model output.

The company also said it does not believe Gemini was used in the operation. Its report noted that China and North Korea-linked actors have shown interest in AI-assisted vulnerability research, including prompt-based security testing and large-scale analysis of known flaws.

Crypto security risks widen

The warning adds to rising concern over AI tools in crypto security. Separate reports have tracked OpenClaw-related phishing, where attackers used cloned websites and malicious wallet prompts to target developers and drain crypto wallets.

Other security coverage has also warned that AI agents can create new weak points when they process outside content, connect to third-party tools, or act without enough human approval. Those risks are more serious when agents can access wallets, private files, browser data, or account credentials.

Google said threat actors are also testing AI for malware support, defense evasion, information operations, and access to AI systems. It named malware families such as PROMPTFLUX, HONESTCUE, and CANFAIL as examples of tools using LLMs for obfuscation or decoy code.

Bitcoin Ordinals explorer Ord.io will shut down on June 1, marking a fresh setback for the Bitcoin inscription market. 

The platform launched in 2023 and served more than 1 million users, according to the project’s public statement.

Creator Leonidas King said the team could no longer keep the project running. He wrote, “In the end we ran out of money and don’t see a path forward.” The statement points to funding pressure at a time when Ordinals activity has cooled from its 2023 and 2024 highs.

Zap also winds down

Zap, a consumer app linked to the same team, will also stop operations on June 1. The app aimed to let users sign up and buy bitcoin memecoins in under 30 seconds, but the team said it failed to reach the user growth needed to continue.

The platform told users to log in and export their private keys before the shutdown. Reports said users were advised to import those keys into Phantom to keep access to assets. Zap also said users who miss the deadline can still access funds through Privy Home.

Meanwhile, Ord.io said it plans to preserve part of its public history before going offline. The project said it would upload upvotes, replies, and public address profiles to GitHub so future developers can use that data if they build a new explorer.

The team also left the door open for another group to take over the platform. That leaves a possible path for Ord.io to survive, but no buyer or operator had been named at the time of the announcement.

Bitcoin inscription activity cools

The shutdown comes after a sharp boom-and-cool cycle for Bitcoin inscriptions. Ordinals allow users to attach data such as images, text, or code to individual satoshis, creating Bitcoin-native digital collectibles. Earlier explainers described Ordinals as a way to make specific satoshis distinct from others.

Runes later added another wave of activity around fungible tokens on Bitcoin. Earlier market updates showed Runes generated $135 million in fees in its first week after the 2024 halving, but activity dropped soon after. May 2024 data showed only two days in a 12-day period generated more than $1 million in fees.

The broader market has since shown mixed signals. OKX launched an Ordinals Launchpad in late 2024 and said trading volume for Ordinals, Runes, and BRC-20 collections on its platform had risen 50% since November. At the same time, Binance had already halted support for Ordinal assets, showing uneven demand across major platforms.

Ord.io’s closure adds to that split market picture. The protocol remains live on Bitcoin, but consumer apps need users, funding, and steady trading activity to survive. For builders, the next test is whether Bitcoin-native collectibles can move beyond short hype cycles and support products that stay open.

Shares in MARA Holdings fell after the bell on Monday as the Bitcoin miner’s first-quarter losses deepened from a year ago and its revenues missed analyst estimates.

MARA’s earnings released on Monday reported its revenue for the quarter ending March 31 dropped 18% year-on-year to $174.6 million, missing Wall Street expectations of $192.7 million.

The company reported a loss of $1.3 billion for the quarter, widening from its $533.4 million loss from the year-ago quarter. Its earnings per share were a loss of $3.31, compared to estimates of a loss of $2.20.

Shares in MARA Holdings (MARA) fell 3.44% in after-hours trading on Monday to $12.93, erasing gains over the trading day, which ended at a gain of 3.48% to $13.39.

MARA Holdings erased gains after the bell on Monday after the company’s earnings missed expectations. Source: Google Finance

MARA stock has fallen 16% over the last 12 months, but has begun to mount a return this year as it has focused on pivoting to build artificial intelligence data centers.

The company reported its first-quarter losses were largely attributed to unrealized losses in its 38,689 Bitcoin treasury as the cryptocurrency fell 23% during the quarter. MARA said it sold more than 15,100 Bitcoin worth $1.1 billion in the final week of March.

MARA said that Bitcoin mining remains the company’s “operational foundation,” even as it continues expanding into AI and high-performance computing to pursue additional revenue streams. 

MARA is one of several US-based Bitcoin miners that have seen profits turn into losses as challenging mining conditions continue to weigh on the sector. 

Bitcoin is trading more than 35% below its all-time high of $126,080, significantly reducing miner revenues per block, while mining difficulty, a measure of how computationally difficult it is to mine a block, has risen nearly 30% over the past year. 

MARA has also lost ground to competitors, falling from the largest Bitcoin miner by market cap to seventh place as rivals have more aggressively expanded into AI.

Related: Saylor signals another Bitcoin buy after hinting at selling in Q1 earnings call 

MARA’s current AI strategy centers on its partnership with Starwood Capital, aimed at converting Bitcoin mining sites into AI and HPC data centers, and Long Ridge Energy & Power, a gas-fired power plant and data center that it acquired for $1.5 billion in late April.

“Our strategy centers on co-locating new infrastructure with existing Bitcoin mining operations,” MARA said. “This approach creates flexibility: we can generate revenue today through Bitcoin mining while preserving the option to redirect power toward AI and critical IT loads as those opportunities mature on the same sites.” 

MARA added that the Long Ridge Energy & Power acquisition could eventually support 600 megawatts of AI computing capacity and that around 90% of its non-hosted mining capacity could be redeployed for AI and IT compute.

The company said it does not have any plans to purchase additional Bitcoin mining hardware.

Magazine: Bitdeer sells all Bitcoin, Metaplanet rejects misconduct claims: Asia Express

Google’s Threat Intelligence Group says it identified what it believes is the first-ever case of hackers using artificial intelligence to develop a zero-day exploit.

The group said in a Tuesday blog post that it had “observed prominent cyber crime threat actors partnering to plan a mass vulnerability exploitation operation,” using a zero-day vulnerability allowing them to bypass the two-factor authentication of an unnamed “popular open-source, web-based system administration tool.”

The exploit required valid user credentials first, but bypassed the second authentication factor, which is often also used to secure crypto accounts and wallets.

AI has been increasingly used in both cybersecurity and by crypto hackers seeking to carry out exploits or scams. AI company Anthropic claimed last month that its recent AI model, Claude Mythos, found thousands of software vulnerabilities across major systems.

Google said it had “high confidence that the actor likely leveraged an AI model to support the discovery and weaponization of this vulnerability,” as the script for the exploit included a hallucination and a format “highly characteristic” of an AI model’s training data.

The report did not specify the threat actor, but Google said that China and North Korea have “demonstrated significant interest in capitalizing on AI for vulnerability discovery.”

LLMs excel at high-level flaw identification 

Google said the vulnerability did not stem from “common implementation errors” like memory corruption, but a “high-level semantic logic flaw” where the developer hardcoded a trust assumption.

This implies the attackers used a frontier large language model (LLM), as the models excel at identifying high-level flaws and “hardcoded static anomalies,” Google added.

Related: AI agents like OpenClaw could drain crypto wallets via ‘malicious skills’: CertiK

Several malware families, such as PROMPTFLUX, HONESTCUE and CANFAIL also use LLMs for defense evasion, generating decoy or filler code to camouflage malicious logic, Google said. 

LLM vulnerability discovery capabilities compared with other discovery mechanisms. Source: Google

Industrialized LLM abuse is increasing 

LLM access abuse is becoming industrialized as threat actors have built automated pipelines to cycle through premium AI accounts, pool API keys, and bypass safety guardrails at scale — effectively running adversarial operations subsidized by trial account abuse.

“By leveraging anti-detect browsers and account-pooling services, actors are attempting to maintain high-volume, anonymized access to premium LLM tiers, effectively industrializing their adversarial workflows.”

Google concluded that as organizations continue integrating LLMs into production environments, the AI software ecosystem has emerged as a primary target for exploitation.

It observed adversaries increasingly targeting the integrated components that grant AI systems their utility, such as autonomous skills and “third-party data connectors,” but threat actors have yet to achieve breakthrough capabilities to bypass the core security logic of frontier models, it stated. 

Magazine: How AI just dramatically sped up the quantum risk for Bitcoin