Core Scientific has lined up a $500m loan from Morgan Stanley, with an option to double it.

Bitcoin (BTC) mining firm Core Scientific has secured a substantial financing line from Morgan Stanley, marking another sign that large banks see opportunity in the infrastructure underlying digital assets and high-performance computing.

The 364-day facility provides $500m in initial capacity with an option to increase the total to $1b, giving the company sizable firepower to expand and reconfigure its asset base. Core Scientific plans to use the proceeds to acquire and develop real estate, cover construction and development costs, and lock in new energy contracts—steps that support both its bitcoin mining operations and its push into hosting workloads for artificial intelligence and other compute-intensive applications.

The deal highlights how miners with significant power footprints and data center expertise are repositioning themselves as broader infrastructure providers rather than pure-play BTC proxies. By tapping a major institution like Morgan Stanley, Core Scientific is signaling both confidence in its growth trajectory and a willingness to tie its capital structure more closely to traditional credit markets. For the bank, the facility offers exposure to a blend of digital asset-linked cash flows and more conventional data center economics, potentially with collateral in the form of real estate and energy agreements.

Miners, AI and institutional credit

Core Scientific’s financing underscores a trend in which large miners seek to diversify revenue streams by courting AI and cloud clients, leveraging existing sites, cooling solutions and power contracts. As demand for training and inference capacity grows, miners with access to stable, relatively cheap energy are pitching themselves as attractive counterparts for hyperscalers and specialized AI firms. At the same time, they must balance these opportunities with the cyclical nature of bitcoin mining, where profitability can swing sharply with the BTC price and network difficulty.

For institutional lenders and investors, these dynamics create both risk and opportunity. Facilities like the one provided by Morgan Stanley allow banks to structure deals that are secured not just by digital assets but also by hard infrastructure and long-term contracts, potentially making them more palatable within existing risk frameworks. Successful execution could encourage more traditional institutions and platforms such as Coinbase’s institutional arm to deepen their engagement with miners through custody, hedging and capital markets services. As regulatory regimes, including MiCA-style frameworks abroad, bring greater clarity to digital asset activities, miners capable of demonstrating diversified, well-financed business models may find it easier to attract large-scale credit and equity capital.

Key Takeaways

• Oracle is preparing to eliminate thousands of positions throughout various departments, with cuts potentially beginning this month.
• The workforce reduction stems from escalating expenses tied to an ambitious AI data center expansion strategy.
• Certain positions targeted for elimination are those Oracle anticipates automating through AI technology.
• The tech giant intends to secure $45B–$50B in funding during 2026 for its cloud infrastructure development.
• Oracle’s Q3 fiscal 2026 financial results are scheduled for release on Tuesday, March 10.

Over the last year, Oracle has pushed hard into AI infrastructure, securing major partnerships with OpenAI, xAI, and Meta. However, this aggressive expansion strategy now carries significant financial implications — including substantial workforce reductions.

Oracle Corporation, ORCL

According to a Thursday Bloomberg report, Oracle is gearing up to eliminate thousands of positions companywide. These workforce reductions may commence as early as this month.

These planned layoffs represent a more extensive initiative than Oracle’s typical periodic workforce adjustments. The cuts will affect numerous business units, with some specifically targeting positions that management expects artificial intelligence to handle in the future.

Earlier this week, Oracle discreetly initiated a review of vacant positions within its cloud computing division, essentially pausing or halting recruitment efforts in that segment.

The underlying issue involves financial constraints. Oracle has invested enormous sums building the data center infrastructure necessary to fulfill its AI cloud service agreements.

Last December, Oracle disclosed that fiscal 2026 capital expenditures would exceed its initial $35 billion projection by $15 billion — bringing the total to $50 billion.

Subsequently in February, Oracle unveiled its intention to raise between $45 billion and $50 billion throughout 2026 to finance additional cloud infrastructure expansion. This funding strategy encompasses a new at-the-market equity offering valued up to $20 billion alongside mandatory convertible preferred securities.

Mounting Expenses, Growing Investor Anxiety

The capital-raising announcement unsettled investors already concerned about Oracle’s increasing debt obligations. The corporation depleted approximately $10 billion in cash reserves during just the first six months of fiscal 2026.

Oracle’s shares declined over 15% throughout the previous year, and the enterprise has fallen short of Wall Street’s revenue projections in eight out of its most recent ten quarterly reports.

As of May 2025, Oracle maintained a global workforce of approximately 162,000 full-time employees.

Major Partnerships and Client Base

Oracle’s principal cloud computing clients comprise OpenAI, Meta, Nvidia, AMD, TikTok parent company ByteDance, and Elon Musk’s xAI venture. The massive $300 billion OpenAI partnership notably elevated Oracle’s position among top-tier cloud service providers.

However, supporting these high-demand customers demands extensive infrastructure — and that infrastructure comes with hefty price tags.

Oracle currently faces the challenge of maintaining its aggressive growth trajectory while implementing greater financial prudence. The upcoming workforce reductions represent one component of this balancing act.

The company will release its third-quarter fiscal 2026 earnings report on Tuesday, March 10.

Quick Overview

• GPT-5.4, OpenAI’s most advanced professional model, became available on March 5, 2026
• Financial professionals gain access to integrated tools connecting FactSet and Third Bridge platforms
• Direct integration now available within Microsoft Excel and Google Sheets environments
• The release positions OpenAI as a direct competitor to Anthropic’s Claude for Financial Services
• ChatGPT Plus, Team, and Pro members can begin using the model immediately

On Thursday, March 5, 2026, OpenAI introduced GPT-5.4, its newest artificial intelligence model designed specifically with professional applications in mind. The update features specialized capabilities tailored for finance industry users.

The updated system can create spreadsheets, documents, and slide presentations with significantly reduced iterations. It performs web searches to compile data and deliver responses to sophisticated queries.

This release incorporates specialized financial capabilities that establish connections with FactSet Research Systems and Third Bridge data services. These features target professionals conducting financial research and developing investment documentation.

The model now functions directly within Microsoft Excel and Google Sheets applications. OpenAI simultaneously released a ChatGPT add-in specifically designed for Excel.

This launch intensifies competition with Anthropic, which previously introduced Claude for Financial Services. Both organizations are vying for enterprise clients prepared to invest in premium AI solutions.

Performance Metrics and Testing Results

During internal evaluations focusing on investment banking spreadsheet operations, GPT-5.4 achieved an 87.3% success rate, representing a substantial improvement from GPT-5.2’s 68.4% score. When human evaluators compared presentations, they selected GPT-5.4 output 68% of the time over its predecessor.

Using GDPval, a benchmark that evaluates AI performance across 44 different professional roles, GPT-5.4 equaled or exceeded human professional standards in 83% of test cases. The previous GPT-5.2 version achieved 70.9% on identical assessments.

The model recorded 75% accuracy on OSWorld-Verified, a benchmark measuring desktop navigation capabilities using visual inputs and pointer commands. Human participants scored 72.4% on the same evaluation.

OpenAI characterizes this as their most accurate model to date. Incorrect statements occur 33% less frequently than with GPT-5.2.

Access and Cost Structure

GPT-5.4 is becoming accessible to ChatGPT Plus, Team, and Pro members starting today under the designation GPT-5.4 Thinking. The preceding GPT-5.2 Thinking version will remain operational for three additional months before discontinuation on June 5, 2026.

Enterprise and Edu subscription holders can activate early access via administrative controls. GPT-5.4 Pro becomes available for Pro and Enterprise tier subscribers.

API usage costs begin at $2.50 per million input tokens and $15 per million output tokens. These rates exceed GPT-5.2 pricing, which stood at $1.75 and $14 respectively.

The system accommodates up to one million tokens of contextual data in Codex. Regular API calls handle 272,000 tokens, with expanded requests incurring double the standard rate.

According to OpenAI, GPT-5.4 delivers superior token efficiency compared to GPT-5.2, potentially offsetting higher per-token costs through reduced overall usage.

The model can be accessed through Codex, the company’s AI development tool, and via the standard API using the identifier gpt-5.4.

Key Takeaways

• AI-related revenue at Broadcom more than doubled during Q1, reaching $8.4 billion thanks to strong sales of custom AI accelerators and networking solutions.
• CEO Hock Tan forecasted that AI chip revenue will surpass $100 billion annually by 2027.
• First quarter adjusted earnings per share reached $2.05, surpassing analyst expectations of $2.03; total revenue of $19.31 billion exceeded projections.
• Management issued Q2 revenue guidance of approximately $22 billion, significantly higher than the Street’s ~$20.5 billion estimate.
• A fresh $10 billion share repurchase program was unveiled, with supply commitments locked through 2028.

Shares of Broadcom advanced approximately 5% during Thursday’s session following robust first-quarter financial results and an optimistic long-term AI growth outlook presented by CEO Hock Tan.

Broadcom Inc., AVGO

The rally followed Broadcom’s report of adjusted earnings reaching $2.05 per share, narrowly beating the Wall Street consensus of $2.03. Total revenue reached $19.31 billion, marking a 29% increase from the prior year and exceeding analyst expectations of $19.18 billion.

The second-quarter outlook proved particularly impressive. Management projected revenue near $22 billion for the upcoming quarter — substantially above the analyst consensus of $20.5 billion.

Artificial intelligence revenue emerged as the standout metric. The segment more than doubled during the period to reach $8.4 billion, propelled by robust demand for customized AI accelerators and networking hardware.

According to Tan, the customer base has expanded beyond established hyperscale cloud providers. Organizations developing AI agents, automated code generation platforms, and consumer-facing AI applications are increasingly adopting Broadcom’s specialized chip solutions.

The company’s AI semiconductor partnerships include major technology giants such as Alphabet, Meta, OpenAI, and Anthropic.

During the analyst call, Tan expressed confidence that the company has clear “line of sight” to annual AI chip revenue surpassing $100 billion by 2027 — a projection that exceeded even the most bullish Street forecasts.

JPMorgan analysts project the company could generate between $12 billion and $15 billion for each gigawatt of AI infrastructure capacity by 2027. Their revised AI revenue projections “conservatively” reach $120 billion or higher.

Analysts at Goldman Sachs highlighted that Broadcom’s “leadership in AI networking and custom silicon enables the lowest inference cost for its hyperscaler customers.”

Supply Agreements and Profitability

Investor concerns about high-bandwidth memory constraints were prominent heading into earnings. Tan directly addressed these worries, confirming that Broadcom has locked in memory supply and advanced semiconductor wafer capacity extending through 2028.

He also dismissed profitability concerns related to increased AI chip rack shipments. According to Tan, the company has optimized production yields and costs to the point where AI business margins are “fairly consistent” with its broader semiconductor portfolio.

The company is approaching 10 gigawatts of deployed capacity distributed across six major customers — a diversification metric that helped alleviate investor worries about customer concentration.

Capital Returns and Street Sentiment

Complementing the earnings report, Broadcom unveiled a new $10 billion stock repurchase authorization, signaling management confidence in the business trajectory.

Wall Street currently rates the stock as a consensus Strong Buy based on input from 30 analysts — comprising 28 Buy ratings and 2 Hold ratings — with a mean price target of $449.46.

Broadcom’s impressive performance created positive ripple effects across related semiconductor names. Credo Technology shares surged 10% while Amphenol climbed 4%, reflecting investor enthusiasm for copper-based connectivity solutions over optical alternatives in AI server architectures.

Tan indicated that AI chip revenue for the current quarter should reach $10.7 billion, signaling continued growth momentum.

Revolut has applied for a US banking licence to deepen its presence in the market.

Revolut, one of Europe’s largest fintechs with a valuation reported around $75b, has applied to the US Office of the Comptroller of the Currency for a banking licence.

If approved, the charter would give the company direct access to core payment rails including Fedwire and ACH, allowing it to offer a broader array of services such as credit cards, personal loans and expanded deposit products. Until now, Revolut has operated in the US via partnerships and a more limited permissions set, which constrained the speed and scope of its product rollout compared with its European footprint.

The application underscores how intensely the firm views the US as a key strategic market, even as competition from incumbents and other neobanks remains fierce. A banking licence would not only improve Revolut’s economics by reducing reliance on third-party intermediaries, it would also give regulators clearer oversight of its balance sheet, risk management and compliance programs. For users, the result could be a tighter integration of fiat, card, savings and crypto functionality—areas where Revolut has sought to differentiate itself by offering exposure to assets like BTC alongside more traditional services.

Fintech, crypto and regulatory convergence

Revolut’s move comes as the boundaries between fintech, traditional banking and crypto services continue to blur. Many digital-first institutions already provide some combination of crypto trading, stablecoin access and on-chain transfers, often in partnership with exchanges such as Coinbase or through their own limited offerings. Securing a full banking licence would position Revolut to more deeply embed these services within a regulated framework, potentially easing concerns for both users and policymakers about the safety and soundness of hybrid platforms.

For US regulators, granting or denying the application will send an important signal about how open the system is to globally active, crypto-friendly fintechs seeking full bank status. The decision will likely take into account not only Revolut’s financial strength and compliance track record, but also broader debates about innovation, competition and consumer protection. As regulatory regimes like MiCA shape expectations in Europe, a US banking licence could help Revolut harmonize its oversight environment across major markets, giving it a stronger base from which to compete with both incumbent banks and emerging digital challengers.

A Coinbase shareholder filed a derivative lawsuit against several of the crypto exchange’s top executives and board members, alleging they failed in oversight of compliance and disclosures, exposing the company to legal and regulatory fallout.

The complaint was filed Tuesday in the US District Court for the District of New Jersey and was brought by shareholder Kevin Meehan on behalf of Coinbase Global. It cites CEO Brian Armstrong, co-founder Fred Ehrsam, and several current and former directors and senior executives, including chief legal officer Paul Grewal and chief financial officer Alesia Haas.

According to the filing, the defendants allegedly made false or misleading statements between April 2021, when Coinbase went public through a direct listing, and June 2023. The plaintiff argues that these oversight failures ultimately exposed Coinbase to regulatory enforcement actions.

In early 2023, Coinbase reached a $100 million settlement with the New York State Department of Financial Services (DFS) over deficiencies in its anti-money laundering (AML) compliance program. In another instance, the company was hit with a $5 million penalty from New Jersey’s Bureau of Securities related to the listing of unregistered securities.

Related: Trump met Coinbase CEO before slamming banks over crypto bill: Report

Shareholder suit seeks damages, insider profit clawbacks

The lawsuit seeks damages on behalf of Coinbase, along with corporate governance reforms and the clawback of compensation and profits allegedly earned by insiders while the company’s compliance issues persisted.

Because the case is structured as a shareholder derivative action, any financial recovery would go to Coinbase rather than directly to shareholders.

The complaint also calls for a jury trial and accuses the defendants of unjust enrichment, abuse of control and breaches of fiduciary duty tied to what it describes as systemic compliance failures.

Cointelegraph reached out to Coinbase for comment, but had not received a response by publication.

Related: Coinbase opens stock and ETF trading to all US users in multi-asset push

Coinbase faces more lawsuits

In January, a Delaware judge allowed a shareholder lawsuit alleging several Coinbase directors conducted insider trading to move forward, despite an internal investigation that cleared the executives. The case claims that insiders, including Armstrong and board member Marc Andreessen, used nonpublic information to avoid more than $1 billion in losses by selling shares around Coinbase’s 2021 direct listing.