Anthropic has signed a term sheet for a $10 billion funding round that would value the artificial intelligence startup at about $350 billion, CNBC confirmed on Wednesday, underscoring the breakneck pace at which capital is still pouring into the most prominent developers of large language models.

The agreement marks one of the largest private capital raises ever for an artificial intelligence company, underlining both the intensity of investor appetite for AI and the escalating costs of competing at the frontier of the technology.

The round is being led by Coatue and Singapore’s sovereign wealth fund GIC, according to a source familiar with the matter, with the discussions remaining confidential. Anthropic declined to comment, while the Wall Street Journal first reported the deal. If completed, the transaction would cement Anthropic’s position among the world’s most valuable private technology firms, narrowing the valuation gap with rivals such as OpenAI, whose valuation has reportedly climbed to around $500 billion.

Founded in 2021 by former OpenAI research executives, including chief executive Dario Amodei, Anthropic has built its reputation around the Claude family of large language models. The company has consistently positioned itself as a safety-focused alternative in the AI race, arguing that reliability, controllability, and alignment will be decisive differentiators as AI systems become more deeply embedded in businesses, governments, and daily life.

The scale of the proposed funding reflects the brutal economics of cutting-edge AI development. Training state-of-the-art models now requires massive clusters of advanced GPUs, long-term access to cloud infrastructure, and the ability to absorb multibillion-dollar research and operating costs before meaningful profits emerge. Investors backing Anthropic are effectively wagering that only a small number of companies will have the capital, talent, and infrastructure to survive this phase — and that those survivors will command enormous strategic and commercial power.

Anthropic’s deepening ties with major technology players illustrate this dynamic. Amazon has already invested billions of dollars in the company and integrated Claude models into its cloud and consumer offerings, strengthening AWS’s position as a key platform for AI workloads. In November, Microsoft and Nvidia announced plans to invest up to $5 billion and up to $10 billion, respectively, moves that align Anthropic with both a dominant cloud provider and the world’s leading supplier of AI chips.

Those relationships also highlight a broader shift in the AI ecosystem, where model developers, cloud operators, and chipmakers are increasingly intertwined. Nvidia’s involvement signals confidence that Anthropic will be a major long-term customer for high-end GPUs, while Microsoft’s interest reflects its strategy of backing multiple AI players even as it remains closely tied to OpenAI.

Product momentum remains central to Anthropic’s pitch. Late last year, the company released Claude Sonnet 4.5, Claude Haiku 4.5, and Claude Opus 4.5, expanding its lineup to address a wide range of use cases, from fast, lightweight tasks to more complex reasoning, coding, and enterprise applications. The releases were aimed at keeping pace with rapid advances from competitors such as Google and OpenAI, which continue to roll out more capable and specialized models.

The funding talks also come at a moment when enterprises are moving from experimentation to deployment. Demand for AI tools that can be embedded into workflows, customer service, software development, and data analysis is accelerating, driving sustained usage and inference costs. That shift strengthens the investment case for leading model providers, but it also raises the stakes: companies that fall behind on performance or scale risk being locked out of lucrative long-term contracts.

At the same time, the eye-watering valuation underscores growing questions about concentration and power in the AI sector. With a handful of firms attracting the vast majority of capital, compute, and talent, policymakers and regulators are increasingly scrutinizing how AI markets are forming and who controls critical infrastructure. While Anthropic has emphasized safety and responsible development, its rising valuation places it firmly among the dominant players shaping the future of the technology.

For Coatue and GIC, the deal represents a bet that AI will continue to reshape the global economy and that Anthropic will emerge as one of the core platforms underpinning that transformation. For Anthropic, the $10 billion raise would provide a formidable war chest to expand computing capacity, attract top researchers, and accelerate partnerships, while signaling to customers and rivals alike that it intends to remain at the very top of the AI race.

Although the transaction has not yet closed and terms could still change, the signed term sheet alone sends a clear message: investor confidence in elite AI developers remains strong, valuations continue to surge, and the battle to define the next era of computing is far from over.

U.S. Treasury yields ticked modestly higher on Thursday as investors positioned ahead of crucial employment data and continued to monitor geopolitical developments that could influence risk sentiment and monetary policy expectations.

The benchmark 10-year Treasury yield rose more than 2 basis points to 4.163%, while the 2-year note increased by less than a basis point to 3.478%. The yield on the 30-year bond climbed about 3 basis points to 4.845%. Because bond prices and yields move in opposite directions, these upticks suggest some selling pressure in long-dated government debt.

Traders have turned their focus toward a slate of economic indicators expected to shape views on the Federal Reserve’s next policy decisions. Thursday’s calendar includes weekly initial jobless claims, and markets are looking ahead to Friday’s U.S. Bureau of Labor Statistics nonfarm payrolls report, due at 8:30 a.m. ET. The payrolls release will be the first on-time jobs report since the 43-day U.S. government shutdown last year, which disrupted data collection and delayed several key releases. The Bureau of Labor Statistics has cautioned that data collected around the shutdown may carry higher margins of error.

Economists surveyed by Dow Jones forecast the U.S. economy added around 73,000 jobs in December, up from 64,000 in November, and expect the unemployment rate to edge down to 4.5%. While headline figures remain well below historical norms, they would mark a modest improvement in the context of the slowing labor market momentum observed in recent months.

Market participants are also eyeing weekly jobless claims, which provide a more immediate, though volatile, snapshot of the labor market. These figures are often watched for early signs of weakening or strengthening that precede the monthly payrolls. Recent data showed unexpected volatility in claims, with economists noting that seasonal adjustment challenges and holiday distortions can obscure trends.

Implications for the Federal Reserve

Market pricing reflects expectations that the Federal Reserve may ease policy this year, with traders anticipating around 61 basis points of rate cuts, according to data compiled by LSEG. The pace of expected cuts has been shaped by slowing employment gains and rising unemployment — data points that could influence the Fed’s assessment of economic conditions at its next policy meeting.

Ian Lyngen, head of U.S. Rate Strategy at BMO Capital Markets, emphasized the importance of employment figures in this context, noting that even with seasonal distortions, the approaching payrolls report will provide valuable insight into labor market dynamics as 2026 begins. Further clarity on job growth could tilt expectations on the timing and magnitude of future rate moves.

Labor Market Trends and Data Challenges

The broader employment picture has shown signs of weakening, with job openings declining more than expected in November and ADP private payroll data pointing to softer-than-expected hiring. These trends suggest that labor demand has cooled relative to earlier in 2025, amid both structural and cyclical pressures.

Data disruptions from last year’s shutdown — which forced the Bureau of Labor Statistics to skip publishing an October unemployment rate — add complexity to interpreting recent figures and assessing the true health of the U.S. jobs market. The November unemployment rate was reported at 4.6%, a four-year high, partly influenced by the prior data gap.

Beyond domestic data, investors are monitoring geopolitical developments for potential risk-off impacts. Continuing uncertainty in Venezuela and high-profile U.S. discussions around acquiring Greenland have kept geopolitical risk premiums elevated in some asset classes, including bonds. Such tensions can drive demand for safe-haven securities like Treasuries, even as yields adjust to economic data.

Movements in Treasury yields reflect a balance between economic indicators, policy expectations, and broader global capital flows. Slight rises in yields can signal modest selling pressure or repositioning, as traders hedge rate cut expectations against signs of persistent inflation or resilient growth. The 10-year yield’s relative stability near current levels also suggests that markets are not anticipating abrupt policy shifts ahead of the jobs reports.

Longer-term yields, such as the 30-year, are sensitive not only to Fed policy forecasts but also to inflation expectations and fiscal outlook. A steepening yield curve — where long-term rates rise faster than short-term ones — can reflect market expectations for future growth or shifting risk premiums.

Investor Positioning Ahead of Data

With major economic releases due this week, investors are likely to remain cautious in their trading strategies. Much of the yield movement reflects positioning ahead of Friday’s nonfarm payrolls, which will likely be central to assessing the pace of labor market softening and its implications for Federal Reserve policy.

In markets where every basis point matters, even marginal moves in Treasury yields can have outsized effects on related financial instruments, including mortgage rates, corporate borrowing costs, and equity valuations. How the labor market data ultimately compares with expectations will be a key driver of market direction in the coming days.

President Donald Trump on Wednesday dramatically raised the stakes in his confrontation with the U.S. defense industry, warning that major military contractors will be barred from issuing dividends or conducting stock buybacks until they significantly accelerate weapons production, expand manufacturing capacity, and fix what he described as persistent failures in equipment maintenance and delivery.

In a lengthy and unusually blunt post on his Truth Social platform, Trump accused defense companies of putting shareholder returns and executive compensation ahead of national security priorities. He argued that capital currently being paid out to investors should instead be redirected into new factories, machinery, and supply chains capable of meeting rising military demand.

“Defense Companies are not producing our Great Military Equipment rapidly enough and, once produced, not maintaining it properly or quickly,” Trump wrote.

He added that “massive” dividends and share repurchases were taking place “at the expense and detriment of investing in Plants and Equipment.”

The president also took aim at executive pay, calling compensation packages at major contractors “exorbitant and unjustifiable.” Until companies invest in new production plants, Trump said, “no Executive should be allowed to make in excess of $5 Million Dollars,” effectively proposing a cap on top management pay linked directly to industrial expansion.

Markets reacted swiftly to the remarks, underscoring investor unease about political intervention in corporate financial decisions. Shares of General Dynamics, Lockheed Martin, and Northrop Grumman each fell about 3% following Trump’s post, wiping billions of dollars off their combined market capitalization.

Trump later singled out Raytheon, accusing the company of being “the least responsive to the needs of the Department of War, the slowest in increasing their volume, and the most aggressive spending on their Shareholders rather than the needs and demands of the United States Military.” He warned that the Pentagon would sever business ties with Raytheon unless it sharply increases investment in plants and equipment, and said that “under no circumstances” should the company carry out any further stock buybacks in the interim.

That warning triggered another sell-off. Shares of RTX, Raytheon’s parent company, slid an additional 2% in after-hours trading after closing down 2.5%. RTX is a cornerstone of the U.S. defense industrial base, manufacturing advanced air-to-air missiles and key components for the F-35 fighter jet, making it deeply embedded in both U.S. and allied military programmes.

Trump framed his intervention as a reallocation of existing resources rather than a call for more government spending. Military equipment, he said, “must be built now with the Dividends, Stock Buybacks, and Over Compensation of Executives, rather than borrowing from Financial Institutions, or getting the money from your Government.” He ended the post with a stark warning to the industry to “BEWARE,” signaling that further action could follow if companies fail to comply.

The legal and practical impact of Trump’s announcement remains unclear. It was not immediately evident whether the restrictions he outlined would be enforced through executive orders, changes to Pentagon contracting rules, or legislation. The White House did not immediately respond to requests for clarification, leaving defense firms and investors uncertain about the scope, timing, and enforceability of the proposed measures.

The comments come at a sensitive moment for the defense sector. U.S. and allied militaries have been drawing down stockpiles of missiles, ammunition, and other equipment amid ongoing global conflicts, exposing limits in production capacity built for peacetime demand. Contractors have repeatedly pointed to supply chain disruptions, shortages of skilled labor, and the long lead times required for specialized components as obstacles to rapidly scaling output.

Trump’s remarks suggest growing impatience with those explanations and a willingness to apply direct pressure on contractors’ balance sheets. His focus on buybacks and dividends taps into a broader debate in Washington over whether defense companies have prioritized financial engineering over long-term industrial investment.

Recent disclosures highlight the scale of shareholder payouts in the sector. Northrop Grumman spent $1.17 billion on stock buybacks in the nine months ended September 30 and paid $964 million in dividends over the same period. Lockheed Martin repurchased $2.25 billion of its own shares in the nine months ended September 28, alongside $2.33 billion in dividend payments.

While such payouts are common across corporate America, Trump’s intervention introduces a new layer of political risk for an industry that has historically enjoyed stable, if closely scrutinized, relations with Washington. Any sustained restrictions on dividends, buybacks, or executive pay could force defense companies to rethink their capital allocation strategies, potentially shifting more cash toward factories, tooling, and workforce expansion.

The episode, for investors, raises questions about how defense stocks should be valued in an environment where shareholder returns could be subordinated to strategic and political priorities. For the companies themselves, Trump’s warning signals that production speed and industrial capacity may now carry consequences that extend well beyond contract awards, reaching deep into boardroom decisions on pay and capital returns.

However, the sharp market reaction suggests it is being taken seriously, and it underscores a clear message from the White House: faster weapons production and expanded manufacturing capacity are no longer just contractual expectations, but conditions that could determine how freely defense companies are allowed to reward their shareholders and executives.