Automaker General Motors on Tuesday announced a new partnership with defense company Lockheed Martin to scale manufacturing and expand production capabilities.
The deal was facilitated by the U.S. Department of Defense, according to Bruce Brown, GM’s vice president of strategy at GM Defense, and will focus on munitions.
“What makes this moment especially important is that the country needs more than great technology. It also needs the capacity to build, scale and deliver reliably,” Brown said on a call with reporters. “This is where GM can help. Across our company, we bring deep experience in advanced engineering, digital development, supply chain discipline and manufacturing at scale.”
Lockheed Chief Operating Officer Frank St. John said it was too early to say what projects it would invest in with GM Defense.
Executives from both companies said on the call that the collaboration will allow for more growth at a time when the country is ramping up its production of defense parts.
“Together, we will explore opportunities across three important areas: improving production readiness and scalable manufacturing environments; strengthening supply chains and identifying ways to increase resilience; and applying advanced manufacturing and design approaches [that] can help improve efficiency and accelerate delivery,” St. John said.
Lockheed Martin is investing $9 billion through 2030 to modernize 20 of its facilities and supply bases, St. John added. GM said it will spend $7 billion on research and development in the U.S., according to Brown.
The executives said the partnership will be focused on “high-rate manufacturing” at scale and expanding production capacity. They added that the collaboration is still in early stages and that they need to further define what the potential for future contracts may be. They are working under a memorandum of understanding.
The automaker built tanks for the country during World War II. Its GM Defense unit is one of the company’s newer but fast-growing business segments, reestablished in 2017 with customers including the U.S. Army, Secret Service and NASA.
“America is stronger when two companies with deep manufacturing roots come together to help expand speed, scale and resilience in the defense industrial base. That is why Lockheed Martin and GM are announcing this collaboration,” Brown said on the call.
The partnership comes as President Donald Trump has been pushing for more American manufacturing to bring more production and reshoring into the country. The U.S. has also seen its defense stockpiles fall because of the wars in Ukraine and Iran.
The White House has held discussions with Ford and GM about better supporting the country’s defense industry.
— CNBC’s Michael Wayland contributed to this report.

As OpenAI files SEC paperwork ahead of an expected initial public stock offering, newly leaked financial documents show a company with quickly growing revenues that are currently being overwhelmed by even larger expenses.
The audited financial statements, obtained by independent journalist Ed Zitron, show OpenAI’s reported revenue growing from $3.7 billion in 2024 to $13.07 billion in 2025. The Financial Times, which reviewed the same documents, writes that the company’s monthly revenues had grown to nearly $2 billion by the end of 2025, suggesting that its ongoing revenue rates continued to grow throughout the year.
But the company’s fast-growing revenues are still dwarfed by its even more significant expenses. OpenAI’s total revenues in both of the last two years were outpaced by research and development alone, which grew from a $7.81 billion line item in 2024 to a massive $19.18 billion cost in 2025. Those numbers seem to reflect the significant costs OpenAI incurred in training new models and include $10.59 billion in R&D costs paid to Microsoft alone in 2025.
On top of that, OpenAI’s “cost of revenue” (i.e., the money spent producing and distributing the product) increased from $2.65 billion in 2024 to $7.5 billion in 2025. This cost line likely reflects the significant compute costs incurred at “inference time” as the company’s models respond to a growing number of user prompts. Costs associated with sales and marketing also grew from $1.11 billion in 2024 to $5.73 billion in 2025.
All told, OpenAI’s day-to-day “loss from operations” increased from $8.78 billion in 2024 to $20.92 billion in 2025, a concerning direction for a company that is telling investors it hopes to be profitable by 2030. But measured as a percentage of revenues, the company’s operating losses slightly improved year to year, from 237 percent in 2024 to 160 percent in 2025.
Gotta spend money to make money
Operating numbers aside, OpenAI’s headline “net loss” number of just over $5 billion in 2024 ballooned to nearly $39 billion in 2025. But the 2025 number includes a significant accounting charge related to investor valuations that shifted amid the company’s 2025 conversion to a for-profit structure. The Financial Times cites “a person familiar with the matter” in reporting that this non-recurring charge was approximately $30 billion and that OpenAI’s 2025 net loss amounted to a more reasonable-looking $8 billion without it.
As OpenAI tries to shift all these losses to eventual profits, it will have to start reining in its costs, especially the massive (and growing) R&D costs associated with model training. It will also have to deal with enterprise customers that are beginning to balk at token-based pricing and starting to demand a measurable return on investment for their AI spending. And on the subscription side, pressure from rival Anthropic may force the company to lower prices, which could further increase operating losses in the near term.
OpenAI shut down its Sora video generation model in March. Around the same time, OpenAI CEO of Applications Fidji Simo told employees that the company would be cutting back on “side quests” and focusing on its core coding and business users.
In March, OpenAI raised $122 billion of financing in a funding round that valued the company at $852 billion. The company reports over 900 million weekly active users of ChatGPT, though only about 50 million of those are paid subscribers.

By this time next year, Fox Corporation CEO Lachlan Murdoch intends to have added Roku to his already expansive media empire. Should the acquisition go through, Fox will gain control of Roku’s modest library of original programming, and the newly combined company will become “the third-largest player in U.S. television” in terms of viewing share. But the most significant thing that Murdoch buys with this deal is direct access to the 100 million households that make up Roku’s user base.
It’s precisely because Roku and Fox are very different kinds of companies that a merger between the two could have a major impact on our modern media landscape. Murdoch has a vested interest in putting a positive spin on the situation — the platform will remain “open” and “partner friendly,” per the announcement — but it’s very much another instance of corporate consolidation that will reshape what consumers watch, perhaps without them ever realizing it.
Fox’s desire to acquire Roku makes a lot of sense when you recall how long Murdoch has been trying to establish a solid presence in the streaming wars. In 2020, Fox bought the free, ad-supported streaming (FAST) platform Tubi, and last year, the company launched its own dedicated subscription service focused on news and live sports. Following news of the deal, Murdoch told investors that he plans to keep Fox and Roku running as two separate entities. The Roku Channel (Roku’s signature FAST service) might not wind up being folded into Tubi, but Murdoch has designs to turn Roku into a doorway to all things Fox.
Both Murdoch and Roku founder, chairman, and CEO Anthony Wood have made it clear that Fox-branded content will be more prominently featured on Roku’s homepage after the acquisition. Wood told investors that “promoting Fox-owned and operated properties on the Roku homescreen” is a key component of the companies’ plan to increase profits. Murdoch added that he expects Roku’s viewership to grow in the US specifically because of Fox content’s presence on the service. Films and series produced by other companies won’t disappear entirely from the platform because the sheer breadth of Roku’s current library is a core part of its appeal to consumers. But it feels more than likely that Fox Sports and Fox News will be some of the very first things people see whenever they interact with Roku’s tech.
The beauty of Roku as it currently exists is its overall simplicity compared to its competitors. When you open the Roku app, it presents you with a small selection of top picks, a list of other streaming services you can access through Roku, and one large ad — all of which makes it feel like the interface is gently trying to help you find things to watch. But during their investor call, Murdoch and Wood stressed that they want to generate more ad revenue from Roku’s homescreen by featuring more of Fox’s content. And that kind of push could effectively turn Roku into a place that feels more attuned to its owner’s conservative politics.
Roku’s value lies in its infrastructure and consumer viewing data
Post-merger, Roku would become the latest example of a moneyed Trump ally buying control of a massive part of the media ecosystem. It took less than a year for Paramount and its associated properties like CBS News and 60 Minutes to be remade in David Ellison’s image. And he obviously intends to do the same with Warner Bros. Discovery when his bid to acquire that studio ultimately closes. Fox’s bid to buy Roku for $22 billion still needs to receive regulatory approval, but chances of the deal being cleared feel high given the Department of Justice’s recent receptiveness to massive media megamergers involving the president’s friends.
Murdoch is trying to do something different with Roku, whose value lies in its infrastructure and consumer viewing data as opposed to the films and series it makes. Murdoch can’t exactly strong-arm competitors like Apple and Amazon into putting out programming that aligns with his political views. And removing other companies’ subscription services from Roku would push consumers away to alternative streaming devices. Instead, it seems as if Murdoch wants to use Roku’s digital real estate to expose Fox’s content to as large an audience as possible. It’s hard to imagine every single Roku user suddenly becoming a Fox News obsessive, but more people will watch it if the platform makes that programming more easily accessible.
This particular scenario — millions of viewers casually consuming Fox content simply because it’s been served up to them — feels plausible because of how thoroughly Roku has already cemented itself in the streaming space. People didn’t buy their Roku-enabled televisions thinking that they were inviting the Fox Corporation into their home. But that’s exactly how all of this is going to shake out if Murdoch gets his way. Given Murdoch’s ties to Trump and this administration’s commitment to pushing the media toward the right, it seems like Roku will soon be a Fox company. And while consumers might not be able to do anything to prevent the deal, there are still other streaming platforms they can jump to.

It appears using AI as a cover story for cutting jobs is fast falling out of fashion.
Unlike many of his tech industry peers who have cut thousands of jobs this year citing the need to restructure their teams to make the most of AI, Robinhood’s CEO Vlad Tenev conspicuously made no mention of AI in his note to employees announcing that the company is letting go 10% of its full-time employees, or about 290 people.
Nor did the company’s regulatory filing announcing the move, which instead framed the cuts as a restructuring exercise.
Still, Tenev did say the company would use “frontier technologies to push our execution even further,” which sounds like a conscious effort to avoid even naming AI. Which isn’t surprising: Sentiment against AI and related infrastructure projects has been trending lower even as a small minority of tech executives make ridiculous bank.
But Tenev did add to the ongoing narrative that it’s now necessary for companies to operate with smaller teams and “flatter organizational structures,” writing: “We ⁠cannot default to operating as a heavily-layered organization. We must be a lean, hyper-focused team where every single individual is empowered to make a massive impact.”
We’ve seen companies of various stripes, from Amazon, Block, Coinbase, GitLab, and Intuit employing similar language in their layoff announcements, indicating that large teams, bureaucracy, and siloed departments are now seen as undesirable line items at a time when AI tools promise to significantly improve productivity.
Some even think it’s a tacit allusion to the fact that tech companies over-hired following the COVID-19 pandemic, and are now scaling back as expenses begin to pile up — especially those associated with massive AI usage.
Regardless, these companies are doing quite well. Tech stocks have surged broadly, spurred by record revenues, improving profit margins (GitLab reported 88% gross margin last month), skyrocketing demand for cloud services, and the belief that the billions being poured into data center projects will produce returns that are orders of magnitudes higher.
Robinhood itself reported a 15% improvement in first-quarter revenue in April, and the company said its second quarter is looking better thanks to rising prediction market fees, subscription revenue, and strong equity and option-trading volumes as markets stabilize.
The company said on Tuesday it is also closing “a small number” of open roles, and that it would incur about $28 million in costs related to the cuts.

Michael Wayland | CNBC
After growing to become one of the largest used car retailers in the U.S., Carvana is expanding into the new vehicle market.
The company has quietly purchased seven new vehicle franchises since last year that primarily sell Stellantis ‘ Chrysler, Dodge, Jeep and Ram brands, including a location in Arizona that has become the automaker’s largest volume store in the U.S.
Dealers and industry experts said they believe the move could significantly disrupt, if not reshape, the century-old new vehicle franchised dealer system.
“Carvana entering the new vehicle franchise business may be one of the most disruptive forces that auto retailing has seen in the U.S. market in decades,” John Murphy, a longtime Wall Street analyst and automotive consultant, told CNBC.
The U.S. franchised dealership system — which includes 16,990 retailers that topped $1.3 trillion in sales last year, according to the National Automobile Dealers Association — has historically been reluctant to change. However, dealers have grown more adaptable in recent years as a means of survival, including during the pandemic and with the rise of publicly traded dealership groups.
watch now
Carvana’s first new car dealership for Stellantis in Casa Grande, Arizona, has grown quickly. It sold more than 700 new vehicles last month, according to Stellantis figures shared with dealers and provided to CNBC.
That made it the bestselling store nationally and compares with an average of roughly 30 to 50 monthly sales the store was doing before Carvana purchasing it early last year, as first reported by The Wall Street Journal.
Carvana and its CEO, Ernie Garcia, have declined to comment about the franchised stores or details of the businesses ahead of a media event this week at which the retailer is expected to disclose its plans.
Carvana: From vending machines to online used car leader
Carvana’s locations, many of which feature its signature large car vending machines, have historically acted as delivery and drop-off points where customers can pick up vehicles they purchased online or turn in a vehicle they sell to the company. And up until last year, those vehicles had been used cars, trucks and SUVs that were largely bought from auctions and individual consumers.
Adding the new vehicle business not only provides additional revenue for the company, it opens up other avenues for Carvana to more easily purchase used vehicles from their new vehicle customers and through exclusive auctions only open to franchised dealers.
“That is a significant game changer in the secondary market,” Murphy said regarding the private auctions. “If that expands to other brands, that is going to be an advantage.”
Jeff Greenberg | Universal Images Group | Getty Images
It also helps Carvana better capitalize on the complete lifecycle of a vehicle. The dealership model is comprised of four main areas of growth: new, used, parts and service, and finance and insurance.
Carvana has previously covered used sales and F&I, including selling consumer auto loans it originates to institutional investors and partner banks, such as Ally Financial , to maintain liquidity. Adding the new franchises is expected to bring Carvana into the other areas as well.
“After stabilizing their core business, I think they realized, by looking at the franchise model, that there was a significant amount of revenue and gross profit opportunity that their business model didn’t even contemplate,” said Brian Gordon, president of dealer advisor and broker Dave Cantin Group.
Dealers adapt or ‘be irrelevant’
Despite Carvana’s current status, which includes a market cap of more than $70 billion, significantly higher than that of Stellantis, there are challenges to selling new cars compared with used.
Unlike used vehicles, which Carvana has specialized in selling online, the sales of new vehicles are more regulated state by state. The franchised owners also act as a business partner to most automakers operating in the U.S.
In some states, such as Michigan, the only way to legally purchase a new vehicle is through a franchised dealer — something direct-to-consumer companies such as Tesla and Rivian have battled with varying results.
An annual study by Cox Automotive, which supports franchised auto dealers, found that most buyers don’t want an all-online purchase or a fully in-person transaction. They want a blend of online convenience with in-store interaction.
Franchised dealers also must adhere to far more regulations and rules from the automakers. They range from showroom layouts and what brands they can sell at certain stores to automaker-defined allocations of vehicles and service and repair requirements, which Carvana does not currently offer for customers.
Not all are mandates, but many automakers incentivize retailers through vehicle allocation as well as financial incentives for offering such services and meeting their requirements.
Carvana is already operating a bit differently though than most dealers, as Stellantis has approved it as a certified website provider for the automaker, which means it doesn’t need to go through an approved third-party company, according to four people familiar with the decision, who requested anonymity to speak about matters that have not been made public.
“It’s bred out of desperation,” said a Stellantis dealer who asked for anonymity to be able to speak freely about the automaker, which has drastically lost U.S. market share in recent years. “It’s given Carvana an opportunity to come into the new car space.”
Stellantis, in an statement to CNBC, said Carvana operates as a “corporate owner” of its brands, similar to other large publicly traded companies such as Lithia and AutoNation .
“We apply the same consistent standards and criteria to all dealer partners, and any organization that meets our qualifications is eligible to operate as a franchisee,” the company said, adding that Stellantis “certifies tools and services that will enhance our program and be beneficial to our network. All certified providers must complete a rigorous onboarding process and meet program standards and requirement.”
Carvana’s foray into new vehicles and its rapid growth have been a discussion between Stellantis’ current dealers and the company, according to Stellantis National Dealer Council Chairman Sean Hogan.
He said competition is always good for the consumer, which is why the franchised dealer model was created, but there are a lot of outstanding questions about Carvana’s new vehicle strategy.
“I’m curious to see what their strategy is and, in the long run, I think competition is good. So, if they’re doing something better than we are, then we will need to adapt, or we’re going to be irrelevant,” said Hogan, vice president of Sierra Auto Group in California.
In JD Power’s annual U.S. Sales Satisfaction Index for franchised dealers that ranks purchase experiences, three out of four of Stellantis’ main brands — Chrysler, Dodge and Ram — were under the industry average.
An Amazon of used and new vehicles?
Although Stellantis said it is treating it like other dealers, Carvana is not a traditional auto retailer like other large publicly traded dealers such as Lithia or AutoNation. It almost exclusively operates online, with a vast network of physical facilities supporting it.
Carvana has built a nationwide logistics and processing company for vehicles similar to Amazon and its back-end operations for processing and shipping consumer goods.
“They have a pre-built out infrastructure, digitally, physically, logistically, that probably gives them an advantage over those big, multibranded public companies,” said Larry Dominique, a longtime automotive executive turned industry consultant.
The business concept of Carvana is simple: buy and sell used cars. But the process behind it has proven to be complicated, labor-intensive and expensive.
Michael Wayland / CNBC
Carvana puts each vehicle it intends to sell through a lengthy inspection, repair and sale preparation process. It ranges from fixing scratches, dents and other imperfections to working on engine and powertrain components. There are also significant logistical costs and processes for delivering vehicles to consumers’ homes.
The other new vehicle Stellantis franchises for Carvana are in Sacramento and San Diego, California; Dallas; Atlanta; Cleveland; and Boston. The new dealerships are in addition to more than 100 other Carvana locations, mainly consisting of vending machines and processing centers.
While large dealers have stores across the country that they can utilize for used and new vehicle inventories, they have traditionally sold regionally to avoid additional shipping costs as well as sales and registration complexities due to selling across state lines.
“Carvana is showing the franchise dealer community how the power of digital can be applied to make a future direction retail model,” Dominique said. “There’s nothing stopping any dealer in the United States from doing that today.”
The company’s vending machine locations do not have parts and service departments, like traditional franchised dealers have, which represent significant profits and customer touch points. That’s one of the main questions surrounding Carvana’s plans: Will it expand into parts and services or leave that for current dealers?
“If they’re going to just be an outlet for new cars, then does that change the dynamic of the dealership model? Who’s going to be responsible for taking care of the customer after the sale?” Hogan said.
Murphy said he believes Carvana may be able to use locations of Adesa, an auction company it purchased in 2022, in addition to the new dealer franchises to potentially service its vehicles.
Carvana has reported it has the capacity to recondition approximately 1.5 million vehicles per year. That compares with its sales of less than 600,000 vehicles last year.
“They do have tremendous capacity to recondition, potentially significantly ramp up their service capability in a way that is not present in other large consolidators,” Murphy said. “I think that problem potentially gets cured.”