Two top Democratic lawmakers are pushing President Donald Trump for answers on the mountain of stock trades made in his name since he returned to the White House.

In a 17-page letter sent Wednesday and seen by CNBC, Sen. Elizabeth Warren of Massachusetts and Rep. Robert Garcia of California asked Trump to reveal the money managers he says are in charge of all his investment decisions.

They also pressed him to disclose his involvement in or awareness of numerous individual stock trades that have raised concerns due to their potential overlap with government policy, or their proximity to market-moving statements made by Trump himself.

“The appearance of numerous conflicts of interest from a President trading in individual stocks – including the potential for insider trading, market manipulation, and policy decisions that benefit the President’s stock portfolio rather than the public interest – undermines public trust in government,” Warren and Garcia wrote in the letter.

“The American people deserve to know that the President is not using his office to increase the value of his stock portfolios, rather than working to bring down the cost of groceries,” they wrote.

The letter is the latest sign that Democrats are gearing up to aggressively investigate Trump if they win control of one or both chambers of Congress in the November midterms. Garcia is ranking member of the House Committee on Oversight and Government Reform.

The Democrats wrote that they sought the responses from Trump in order to inform their work on potential legislation to ban presidents from owning individual stocks. U.S. presidents are currently allowed to trade securities, but legally required to report their transactions.

“The American people deserve to know if the president is working for them or for his own stock portfolio,” Warren said in a statement to CNBC. “It’s past time we ban Members of Congress, the Vice President, and the President from owning and trading individual stocks.”

Trump has defended the staggering number of trades made in his name — more than 21,000 in 2025 alone, according to an analysis of his 927-page annual financial disclosure — by insisting that he has no contact with the institutions that run his money.

CNBC previously linked JPMorgan Chase, Charles Schwab, UBS and Stephens Inc. to at least four of the eight investment accounts listed in Trump’s 2025 disclosure. A full list has not been disclosed, and details about how outside firms handle Trump’s portfolio — and how and when they make decisions about what to buy and sell — remain unclear.

“We have funds that run my money,” Trump said in July. “They invest my money, and I don’t talk to them. I never — I don’t even speak to them.”

Wednesday’s letter from Warren and Garcia asked Trump to list…

Some of America’s biggest technology companies are pushing Washington to make open artificial intelligence a national priority as China continues to build its lead in the space.

Their message exposes a gap in U.S. policy: America has a chip strategy for AI. Now it needs an open-model strategy.

Supporters say open models can lower costs, keep sensitive data inside a company’s own systems and prevent businesses from becoming dependent on a handful of AI providers. The models can also give more researchers and security teams the ability to inspect powerful models and find weaknesses.

Unlike ChatGPT or Claude, which customers access through services controlled by OpenAI and Anthropic, open-weight models can be downloaded, customized and run on a company’s own computers. That can make them cheaper and give users more control over their data and technology.

Last week, a coalition led by Nvidia published an open letter arguing that America’s AI leadership will depend on building a strong open ecosystem. Microsoft, Meta and other major technology companies signed it. OpenAI and Google later added their support.

A striking cybersecurity incident days earlier showed why the issue matters.

During an internal test, OpenAI models found a way out of a restricted environment and compromised systems at the AI platform Hugging Face. When Hugging Face tried to investigate, safety controls on closed commercial models blocked them from analyzing the attack. The company instead used an open Chinese model that it could run on its own infrastructure.

A closed American model caused the incident. A Chinese open model helped investigate it.

That episode complicates the argument that closed AI is always safer.

America has restricted China’s access to advanced chips, spent billions to onshore semiconductor manufacturing and encouraged a massive buildout of data centers and power. Meanwhile, Chinese labs like Zhipu and Moonshot AI have released capable open models at a fraction of the cost of leading American systems.

The Trump administration is now reportedly considering restrictions on Chinese models. But a ban would not stop those models from spreading around the world. It could simply leave American developers on the sidelines.

China has a clear incentive to push open-weight AI. Every improvement makes powerful models cheaper and weakens the advantage of American companies that charge a premium for access to closed systems.

OpenAI and Anthropic have the opposite incentive. Their businesses depend on keeping their best models proprietary. Anthropic CEO Dario Amodei has also warned that models anyone can download are harder to monitor and could be modified by bad actors.

Those security concerns are real, but so is the risk of allowing China to become the foundation on which the rest of the world builds.

The way to win an open-source race is by out-building, not with a ban.

The U.S. could support open AI the way it supported chips, by giving universities and startups access to…

Shares of Super Micro Computer jumped 15% after the server maker said Tuesday it now sees higher margins than previously projected for the June quarter, with a surge in new orders.

The company said its gross margin and adjusted gross margin should come in between 15% and 17%, a step up from the range of 8.2% to 8.4% that management provided in May.

The revision is “primarily due to a favorable customer and product mix,” Super Micro said in a preliminary business update.

Demand for servers containing Nvidia graphics processing units that run artificial intelligence models has been surging for Super Micro, as well as rivals Dell and Hewlett Packard Enterprise. Dell stock moved up 5% in extended trading after hours on Tuesday, while HPE gained 4%.

In June, Super Micro CEO Charles Liang wrote on X that he was “proud to co-build another new Gigawatt AI datacenter for @SpaceX and @XAI within a year.” Elon Musk‘s SpaceX Musk’s SpaceX acquired his AI venture, xAI, in an all-stock transaction in February, and is now known as SpaceXAI. SpaceX also owns and operates social network X.

For the June quarter, Super Micro now expects revenue to come in at the low end of its guidance range of $11.0 billion to $12.5 billion, according to Tuesday’s statement. Analysts polled by LSEG were looking for $11.67 billion.

Super Micro said its backlog hit record levels at the end of the 2026 fiscal year, which ended on June 30. It had received over $60 billion in new orders in the fiscal fourth quarter.

“These new orders are expected to be delivered over future quarters,” the company said.

Super Micro expects to hold an earnings call on Aug. 11.

WATCH: Options Action: Super Micro breaks out

U.S. investors are getting a new way to buy into the memory boom.

SK Hynix, the second most valuable company in South Korea, behind only Samsung, is slated to begin trading on the Nasdaq on Friday. The market debut, coinciding with the company’s hefty expansion plans for the U.S., comes after a more than sevenfold increase in its stock price over the past year, a rally that’s lifted SK Hynix’s market cap to about $1 trillion.

Many Americans don’t know the name SK Hynix, but they’ve likely had plenty of exposure to the company’s products. Along with Micron and Samsung, SK Hynix is one of three primary makers of computer memory used in laptops and phones sold by companies such as Apple and Dell.

Memory has long occupied a sleepy corner of the semiconductor market. That’s changed dramatically over the last couple years as surging demand for artificial intelligence has upended the industry, creating severe shortages for the chips and sending prices to all-time highs.

SK Hynix is the leader in the high-performance memory that’s used in AI chips from Nvidia, the world’s most valuable company. While memory components are typically relatively simple chips, often called RAM, that every computer needs to store data, high-bandwidth memory (HBM) is more complicated, as it stacks many layers of traditional memory together. SK Hynix was the first to do it, and analysts project the company will capture more than half the market this year.

In June, Nvidia CEO Jensen Huang visited SK Hynix on a trip to Seoul, as the two companies announced a multiyear partnership. Nvidia is the biggest HBM buyer.

“That advantage has positioned SK Hynix as one of the biggest beneficiaries of the rapid growth in AI infra,” said TrendForce analyst Ellie Wang.

SK Hynix will trade on the Nasdaq under ticker symbol SKHY (initially SKHYV.) The company plans to raise around $29 billion by issuing American depositary receipts (ADRs) in order to fund new factories and equipment, according to a regulatory filing.

Some of that buildout is taking place in the U.S. SK Hynix is constructing its first production facility in the country, scheduled for completion in 2028. The $4 billion plant in West Lafayette, Indiana, will be used for what’s known as advanced packaging, an essential part of the process for producing HBM that involves connecting and stacking individual chips into larger systems that can be used in a computer.

SK Hynix expects to receive up to $458 million in funding from the U.S. CHIPS and Science Act, passed in 2022 as the centerpiece of a plan to build up the domestic chip industry. The company could also receive up to $570 million of loans from the U.S. Commerce Department.

It’s not just HBM that’s boosting SK Hynix. The whole memory industry is experiencing soaring prices due to the AI crunch, which means…

President Donald Trump on Monday morning rang the opening bell of the stock market from the White House Oval Office in a first-of-its-kind joint ceremony with the New York Stock Exchange and the Nasdaq.

The event marking the launch of “Trump Accounts,” the administration’s new tax-advantaged investment vehicles for U.S. kids, was attended by an array of CEOs and other officials who are supporting the new accounts.

They include: Dell chief Michael Dell and his wife, Susan Dell, who together pledged to donate more than $6 billion to the program; Brad Gerstner, Altimeter Capital and CEO and founder of the Invest America charitable foundation, which pushed for the accounts; Intercontinental Exchange CEO Jeffrey Sprecher; Sen. Ted Cruz, R-Texas; Treasury Secretary Scott Bessent; U.S. Securities and Exchange Commission Chair Paul Atkins; and NYSE President Lynn Martin.

“This makes real the promise of the American dream, not for some but for everybody,” Gerstner, a major proponent of the accounts, told CNBC’s “Squawk Box” before the event.

The Trump Accounts, which are available for all children age 18 or younger, include a one-time $1,000 pilot program contribution from the U.S. Treasury Department for babies born from 2025 through 2028.  

Robinhood CEO Vlad Tenev, arriving at the White House before the opening bell, called the accounts potentially “life changing.”

This is developing news. Please check back for updates.

Zhipu‘s GLM 5.2 artificial intelligence model landed last week with the kind of Silicon Valley buzz that followed DeepSeek‘s launch last year.

The Chinese AI startup‘s new open source model has blown past every other open release and now sits within a percentage point of Anthropic‘s Opus 4.8 on one closely watched agentic benchmark, at roughly a fifth of the cost.

Developers are piling in, with OpenRouter token traffic climbing faster than it did after DeepSeek’s V4 launch in April. And unlike DeepSeek, which the market eventually dismissed as a one-off chatbot shock, GLM 5.2 is strong at agentic work like planning, coding, testing, and looping, the kind of work that enterprises are racing to automate.

Companies hit by unexpectedly high AI spend on tokens, the measure of the data processed and generated by AI models, are now asking how to get the most for their money.

The most important metric is becoming intelligence per dollar, making Zhipu’s cheap-but-good-enough model an attractive answer. 

“I’ve been consistently surprised by how quickly the open source has caught up,” Gabe Pereyra, co-founder of Harvey, told CNBC. “GLM 5.2, you’re seeing the first model where it’s really competitive with some of these closed-source frontier models.”

But the real story underneath the price tag is the open source AI moment.

GLM 5.2 is free to download, fine-tune, and run on an enterprise’s own servers, putting pricing pressure on frontier labs at the same time that access looks shaky. Anthropic had to pull its Fable Mythos-class model after an order by the Trump administration, and OpenAI announced Friday that it is limiting its GPT 5.6 models because of a government request.

The federal oversight has made a model that no one can revoke increasingly look like the safer bet.

Elon Musk became the world’s first trillionaire thanks to his stake in SpaceX after the company’s public debut on June 12. Prediction market traders think that Mark Zuckerberg has the best chance of being next, but it’s still a long shot.

Speculators on Kalshi give the Meta CEO a 32% chance of becoming the world’s second trillionaire. His net worth is estimated at just under $200 billion, according to Forbes, which Kalshi uses to determine whether to resolve the contract to “yes” or “no.” That means his net worth would have to quadruple to earn the title.

The contracts on Kalshi related to the question also expire by 2033, meaning if the person listed on the contract doesn’t become the second trillionaire by that point the contract will close. Kalshi’s event contracts related to the question also currently have low volume, with just over $7,500 traded.

Traders on the platform give Nvidia CEO Jensen Huang the next best odds, with 21% chance of obtaining a 13-digit net worth. His current net worth according to Forbes is a little north of $180 billion. 

No one else is seen as having a more than 10% chance of becoming the second trillionaire. Michael Dell, CEO of Dell Technologies, has the third best chances, at 6%. That’s despite his current net worth, $240 billion, being greater than that of Zuckerberg or Huang’s.

Despite the low odds from prediction market traders, more than one trillionaire may be in the pipeline, if previous research is to be believed. An Oxfam report from January 2025 estimated that within a decade there would be five trillionaires. 

Disclosure: CNBC and Kalshi have a commercial relationship that includes customer acquisition and a minority investment.

A version of this article first appeared in CNBC’s Inside Wealth newsletter with Robert Frank, a weekly guide to the high-net-worth investor and consumer. Sign up to receive future editions, straight to your inbox.

Last month, investment firms of the ultra wealthy went all-in on sports and played on multiple fronts.

At the beginning of May, billionaire Tom Dundon’s namesake family office partnered with Apollo’s new sports fund to invest $225 million in Pickleball Inc., the parent company of Major League Pickleball and the PPA Tour. Dundon is already an owner of the Portland Trail Blazers NBA team and the NHL’s Carolina Hurricanes.

As for the major leagues, Michael Dell, as part of an investor group led by Silver Lake’s Egon Durban, bought a 25% stake in the Las Vegas Raiders football team. Dell is also a minority investor in the NBA’s San Antonio Spurs and the Austin Gamblers, a professional bull riding team.

Family offices made 51 direct investments in companies in May, holding steady from April’s deal tally, according to data provided exclusively to CNBC by Fintrx, a private wealth intelligence platform.

A Goldman Sachs survey released last fall found that 25% of family offices have invested in sports or related assets like ticketing or arenas and another quarter are interested in doing so. Asides from love of the game, many investors are drawn to the sector as an inflation hedge.

Student housing mogul David Adelman has used his family office to extend his reach into the sports economy. Adelman is a part-owner of the NBA’s Philadelphia 76ers, English Premier League club Crystal Palace and the New Jersey Devils ice hockey team and an investor in sports merchandise giant Fanatics.

In May, his investment firm Darco Capital co-led a $12 million Series A round for PlayerData along with David Blitzer’s family office Bolt Ventures and venture capital firm Pentland Ventures. The UK startup makes GPS-enabled vests and soccer balls that help athletes track their performance.

Adelman told CNBC that some teams in his portfolio use PlayerData products. Crystal Palace, for instance, uses the vests and smart soccer balls in training its academy players. The technology’s application for athletes at all levels, including youth sports, was part of the startup’s draw, according to Adelman.

“What stood out to me was the ability to take something complex and make it simple, practical, and accessible,” Adelman said.

Hewlett Packard Enterprise shares skyrocketed 30% on Monday after the tech company posted blockbuster second-quarter results that blew away estimates.

Here’s how the company did compared to LSEG estimates:

• Earnings per share: 79 cents adj. vs. 53 cents expected
• Revenue: $10.68 billion vs. $9.79 billion expected

It was the company’s biggest EPS beat since February 2018.

Revenue was up 40% over a year ago.

Overall Cloud & AI revenue came in at $7.71 billion, topping the StreetAccount estimate of $6.87 billion, but it was the company’s server unit that really impressed. Server revenue, which is a sub-division of the cloud and artificial intelligence unit, came in at $5.45 billion, blowing away the $4.66 billion expected by analysts.

The server maker bumped its full-year EPS guidance by a full dollar, projecting fiscal year 2026 EPS of $3.35 to $3.45, up from $2.30 to $2.50. The company said it is now tracking two years ahead of its own long-term financial plan.

CEO Antonio Neri told CNBC’s Kristina Partsinevelos that traditional server bookings are up triple digits, and it is the biggest backlog the company has ever seen.

“Customers continue to invest in modernizing their infrastructure and scaling AI, and our performance shows the strength of our combined networking portfolio,” Neri said in a release announcing the quarterly results.

Neri also told CNBC that industries focused on security are seeing especially rapid acceleration for AI on-premises, instead of in the cloud. HPE is largely focused on providing that service to national labs and enterprises.

Those are “higher margin opportunities” than the neoclouds that competitors like Dell are focused on, according to chip analyst Patrick Moorhead of Moor Insights and Strategy.

“Their bump up has everything to do with increasing profitability for AI, and the fact that they’re hitting their targets a year and a half in advance,” Moorhead said.

The global memory shortage remains a risk, and Neri said he sees costs remaining elevated into 2027.

Net income came in at $624 million, or 44 cents per share, compared to a net loss a year ago. HPE posted a net loss of $1.05 billion, a loss of 82 cents per share, a year ago.

The stock surge comes following the company’s announcement of a new server rack on Monday at the Computex conference in Taiwan. The new server will be powered by Nvidia‘s new Vera central processing units, which are now in full production.

“This is going to be our new major growth driver,” Nvidia CEO Jensen Huang said during his keynote at Computex on Monday. He said millions of the new CPUs are currently being made, and will be available starting in the fall.

The New York Stock Exchange is one customer that plans to use the new Nvidia chips on HPE’s new server, to help it process more than a trillion messages per day.

Neri said it’s a task well suited to his company’s newest ProLiant server because it’s optimized for agentic AI workloads.

“These…