The Supreme Court won’t consider the fate of President Trump’s blanket tariffs until November, but the White House already appears to be making adjustments that could keep as many duties as possible in place if the administration loses. It’s a refocusing of sorts on the tariff authority derived from Section 232 of the Trade Expansion Act of 1962, which allows the president to target sectors of the economy based on national security considerations, in a series of moves already leading to another bout of uncertainty in the business world. Those powers are more legally established than presidential authority under the International Emergency Economic Powers Act of 1977 (IEEPA), a separate law Trump used to impose his “Liberation Day” tariffs six months ago. His use of IEEPA is at the heart of the case now before the Supreme Court. Trump’s pivot to Section 232 can be found on his Truth Social feed, which has been marked by a slew of sector-specific tariff promises on industries ranging from pharmaceuticals to semi-trucks to furniture. More sectors are likely close behind, with long-promised semiconductor tariffs the biggest shoe yet to drop. And of course, trade talks with China are set to heat up before a meeting between Trump and Chinese President Xi at the end of this month. The early lessons so far? Sectors could be in for plenty of whiplash. “The challenge with tariffs is they create a ton of uncertainty,” notes Natasha Sarin, the co-founder of the Budget Lab at Yale. “That isn’t uncertainty that is likely to be resolved by the court,” she added. “Even a ruling against the administration means you’ll likely see them turn to other trade authorities down the road.

‘We don’t know exactly what plan B is yet’

As for the question of when uncertainty may fade, “it’s probably going to be a while,” Stifel Chief Washington policy strategist Brian Gardner told Yahoo Finance this past week. Gardner’s expected timeline is lengthy, given that Trump is in the process of building out his plan B. But “we don’t know exactly what plan B is yet” as well as Trump’s likelihood of using tariffs as a tool of foreign policy for his entire term. Up first is the new Supreme Court term, which begins by law on the first Monday in October. The justices will hear tariff arguments on Nov. 5 and have signaled that a decision could be in hand before the end of the year. That ruling could either uphold these IEEPA powers or invalidate the duties and perhaps even mandate the administration to offer refunds. That’s why Trump and his team have been hard at work building out this insurance policy of sorts around 232 tariffs, which have been used for years by presidents in both parties and are seen as relatively secure from legal challenges. Trump had already used the 232 authority this term to impose tariffs on things like steel, aluminum, and copper. More tariffs under that authority have been promised in recent weeks, ranging from 25% to 100% on goods from pharmaceuticals to semi-trucks to kitchen cabinets and furniture. Some are already in place, and others are coming later this month. The Trump administration also broke with some government shutdown precedents by declaring this past week that ongoing tariff investigations are essential government functions and could continue. The effect there is to keep the wheels turning on active Section 232 investigations, which would then keep the future implementation of duties on goods from critical minerals to semiconductors on track. Areas ranging from robotics to medical devices to personal protective gear could follow after the administration recently opened new investigations in these areas.

A possible lesson from the pharmaceuticals rollout

Meanwhile, a lesson from this past week around pharmaceuticals suggests that the effect of these sector-specific tariffs may be harder to predict. At the very least, they will be subject to a lot more back-and-forth with Trump and his team. As recently as this summer, the president was promising what sounded like broad-based pharmaceutical tariffs, including that companies “are going to be tariffed if they have to bring the pharmaceuticals into the country at a very high rate, like 200%.” That rate then fell to 100%, and more importantly, exception after exception was offered. First, it was that any pharmaceutical company “building” in the United States would be exempt from tariffs. Then, exemptions were extended to any drugs imported from a nation where a trade deal has been struck. Then early this past week, the president announced Pfizer (PFE) is in line for a three-year exemption from tariffs as part of a deal in which the company agreed to slash some of its drug prices. At the event with Pfizer, Commerce Secretary Howard Lutnick suggested that those drug price talks mean exceptions for other companies. “While we are negotiating with these companies, we’re going to let [the talks] play out,” Lutnick said, adding that tariff enforcement for now is “standing by.” It means these tariffs are fully on hold while the administration launches new talks with the industry on bringing their plants to the US, as well as their prices. It has clearly set up an unpredictable situation for drug companies that could stretch for months or even years and be a template for other sector tariff rollouts to come. But it might not be all bad for investors and business leaders. After all, this week’s turmoil saw pharmaceutical stocks have their best week in 16 years.

Tesla (TSLA) on Sunday teased an October 7 event, as investors and analysts await a more affordable model to sustain sales momentum, sending shares of the company 2% higher in premarket trading.

In a nine-second video posted on social media platform X, the Elon Musk-led automaker showed a vehicle with its headlights illuminated in a dark setting. The company hinted at an event scheduled for Tuesday in a separate video that had “10/7” at the end.

Tesla has previously delayed rolling out a lower-cost version of the Model Y in the United States. The company said in June that it had made “first builds” of the vehicle, but would start selling it in the fourth quarter and ramp up output slower than planned.

The stripped-down version is designed to be roughly 20% cheaper to produce than the refreshed Model Y and could scale to about 250,000 units a year in the U.S. by 2026, sources told Reuters earlier this year.

The teaser follows Tesla’s record quarterly deliveries for the three months ended September, driven by a surge in EV purchases ahead of the expiration of the $7,500 U.S. EV tax credit on September 30. However, analysts expect sales to drop off in the coming months without the incentive.

Wall Street expects Tesla’s deliveries to jump next year to 1.85 million vehicles, with the cheaper model accounting for 155,610 units in 2026, according to Visible Alpha estimates.

Tesla faces challenges from an aging model lineup. The company has not introduced a new mass-market vehicle in years, relying heavily on incremental updates to the Model 3 and Model Y to drive sales.

Sales of its last major launch, the Cybertruck, have struggled, with Tesla offering thousands of dollars in discounts for vehicles in inventory in the past few months. A U.S. recall filing in March showed 46,096 Cybertrucks had been built between its introduction in November 2023 and early this year.

The tie between startups and the U.S. government have strengthened in recent years, a shift buoyed by an interest in using AI, automation, space, robotics, and climate tech for defense. And while that has provided another welcome path to capital, the relationship is getting complicated.

A growing share of startups have the U.S. government as customers, or are aiming for permits and defense-related contracts. When the government is operational, that connection can provide a needed boost and revenue to startups. But when the government ceases to function, as it did starting October 1, those close ties can stifle or even halt progress for startups.

This week on Equity, Anthony Ha, Max Zeff, and I (Kirsten Korosec) talk about how a prolonged U.S. government shutdown poses more risk for startups than in the past — not to mention put a damper on an active IPO season. The three of us dug into a few other topics too, including the how AI companies are trying to monetize and the U.S. government’s latest push to take ownership stakes in the tech and industrial sectors.

“This also feels like a reflection of how the startup landscape has changed in say the last decade and especially over the last few years,” Ha said during the Equity podcast, adding the focus was on consumer internet startups for a long time. “Obviously there’s a lot more going on in defense tech, a lot more in deep tech where you maybe need various kinds of regulatory approvals,” he continued. “And so, it feels like much broader swaths of the startup landscape now depend on the government in various ways, in ways that wasn’t necessarily true 10 years ago.”

But it’s not just startups. The Trump Administration has also continued to extend its reach, and ownership, into the tech industry, too.

The Trump Administration has renegotiated yet another federal loan — it’s third in recent months followed by one with Intel and rare earth miner MP Materials — and taken an equity stake as part of the newly hashed out deal.

The U.S. government took a 5% stake in Canadian miner Lithium Americas and another a 5% ownership in a Lithium Americas-GM joint venture to mine lithium in Nevada. The equity stakes will be acquired through no-cost warrants, which are financial instruments that give the government the right to purchase shares at a set price. The new terms came out of a renegotiation with the DOE’s Loan Programs Office of a $2.26 billion loan that was awarded to Lithium Americas under the Biden Administration.

Watch the full episode to hear more about the government’s relationship with startups and tech companies as well as the entertainment industry’s reaction to AI-generated actress Tilly Norwood, and an eye-popping seed round for Periodic Labs.

New PitchBook data illustrates how dramatically AI is dominating startup investment, with 2025 on-track to become the first year when AI accounts for more than half of all VC money invested.

PitchBook reports that VCs have poured $192.7 billion into the industry so far this year, out of a total $366.8 billion, according to Bloomberg. In the most recent quarter, AI accounted for 62.7% of the money invested by U.S. VCs, and for 53.2% of money invested by global firms.

Most of that money is going to marquee names like Anthropic, which announced a $13 billion Series F in September. Meanwhile, the number of startups and venture funds successfully raising money are at their lowest levels in years — PitchBook says that 823 funds have been raised globally so far in 2025, compared to 4,430 in 2022.

PitchBook’s director of research Kyle Sanford told Bloomberg that the market is becoming “bifurcated,” where “you’re in AI, or you’re not” and “you’re a big firm, or you’re not.”

Of all the ways a government shutdown can disrupt the economy, market watchers should be aware of one in particular: an absence of critical new data.

Right alongside furloughs and a pause in operations is the interruption of crucial work product, including the data that central bankers, corporate executives, analysts, and investors rely upon to interpret the economy and make decisions about money.

The timing is significant, and unless an agreement somehow appears, the all-important payroll report for September — scheduled for release on Friday by the Bureau of Labor Statistics — will be delayed, leaving the Fed potentially flying blind into its next interest rate decision.

All of this is especially worrisome because of how data-dependent the central bank has been under the leadership of Chair Jerome Powell, watching each release and carefully analyzing and cataloging it among the “totality” of data. Top-tier releases like the jobs report and both flagship inflation reports put out by the government will leave massive data holes.

But even before the latest threat of a government shutdown, policymakers and economists have signaled a shift toward using a wider variety of data sources, such as private-sector data from ADP’s private payrolls, out on Wednesday morning.

Those data holes — or at least the labor market ones — are getting filled.

In a speech last month, Federal Reserve governor Christopher Waller, who is in the running to become the next chair of the central bank, said he uses ADP figures to assess the labor market, which he sees as continuing to deteriorate. And after US private payroll numbers published by ADP earlier in the summer also appeared to catch the weakening in the labor market before the government revised its own data, Waller’s approach seems particularly canny.

Wall Street economics teams have long used a wide slate of data sources. But last week, Jeffrey Gundlach’s DoubleLine, a prominent bond firm, said that it is using a variety of official and private data sources to assess the economy, precisely because of a decline in the quality of data published by the BLS.

“Confronted by challenges to the accuracy and reliability of BLS data … it is more important than ever to possess robust, multifaceted approaches to economic analysis,” Ryan Kimmel, fixed income allocation strategist at DoubleLine, said in a note last week describing how data collection had narrowed significantly over the years, using modeling to replace actual observations.

In August, after downward revisions to previous jobs data jolted markets, President Trump fired the commissioner of the BLS. The decision prompted even more debate over the future accuracy of the bureau’s widely read employment and inflation reports — which had already been under the microscope given the agency’s hiring freeze.

Still, Wall Street, economists, and investors greatly value government data from the BLS and other departments. And in a year filled with uncertainty, a halt to critical data will only add to the murkiness as the Fed, without its instruments, keeps the plane on autopilot and hopes that’ll be enough.

OpenAI has acquired Roi, an AI-powered personal finance app. In keeping with a recent trend in the AI industry, only the CEO is making the jump.

Chief executive and co-founder Sujith Vishwajith announced the acquisition on Friday, and a source familiar with the matter told TechCrunch he is the only one of Roi’s four-person staff to join OpenAI. Terms of the deal were not disclosed. The company will wind down operations and end its service to customers on October 15.

The Roi deal marks the latest in a string of acqui-hires from OpenAI this year, including Context.ai, Crossing Minds, and Alex.

While it’s not clear whether any of Roi’s technology will transfer over to OpenAI or which unit Vishwajith will join, the acquisition clearly aligns with OpenAI’s bet on personalization and life management as the next layer of AI products. Roi brings a specialized team that has already tried to solve personalization in finance at scale — a challenge whose lessons can be applied more broadly.

New York-based Roi was founded in 2022 and has raised $3.6 million in early-stage funding from investors like Balaji Srinivasan, Spark Capital, Gradient Ventures, and Spacecadet Ventures, according to PitchBook data. Its mission was to aggregate a user’s financial footprint, including stocks, crypto, DeFi, real-estate, and NFTs, into one app that can track funds, provide insights, and help people make trades.

“We started Roi 3 years ago to make investing accessible to everyone by building the most personalized financial experience,” Vishwajith wrote in a post on X. “Along the way we realized personalization isn’t just the future of finance. It’s the future of software.”

Beyond tracking trades, Roi gave users access to a financially savvy AI companion that responded in ways that made sense for them. When signing up, users could personalize Roi by providing information like what they do for a living and how they wanted Roi to respond to them.

In one telling example that Roi posted on X, the sample user wrote: “Talk to me like I’m a Gen-Z kid with brain rot. Use as little words as possible and roast me as much as you want I don’t mind.” In response to a query about the status of the user’s portfolio, Roi replied: “Suje, you got cooked lil bro. Cause of the tariff announcements, you took an L today of $32,459.12…Based on your risk preference this might be an opportunity to buy the dip.”

The exchange highlights the philosophy behind Roi and its co-founder — that software shouldn’t just provide generic answers but should adapt, learn, and communicate in ways that feel personal, human, and most importantly, keep you engaged.

As the Roi team wrote in a blog post: “The products we use every day won’t remain static, predetermined experiences. They’ll become adaptive, deeply personal companions that understand us, learn from us, and evolve with us.”

That vision dovetails with OpenAI’s existing consumer efforts, including Pulse, which generates personalized news and content reports for users as they sleep; the Sora app, a TikTok competitor filled with AI-generated content, including personal cameos from users; and Instant Checkout, a feature that lets users shop and make purchases directly in ChatGPT.

The deal also comes as OpenAI beefs up its consumer applications team, led by former Instacart CEO Fidji Simo. It’s a further signal that OpenAI isn’t just trying to be an API provider, but wants to build its own end-user apps. Roi’s talent and tech could slot right into these apps and help make them more adaptive.

Vishwajith, alongside his co-founder Chip Davis, used to work at Airbnb, where he developed a knack for optimizing user behavior to drive revenue. By his account, a simple change of 25 lines of code led to $10+ million in additional cash.

Being able to bring in meaningful revenue via consumer apps is more important than ever to OpenAI as it continues to burn through billions on data centers and infrastructure to power its models.

US stocks rose on Thursday as the AI trade continued to power a push for fresh records amid new buzz around OpenAI (OPAI.PVT). Meanwhile, investors kept an eye on developments in Washington, weighing the chances of a lengthy US government shutdown.

The tech-heavy Nasdaq Composite (^IXIC) rose 0.4%. The S&P 500 (^GSPC) gained 0.1%. The Dow Jones Industrial Average (^DJI) rose 0.2%. All three major averages posted new records.

Stocks rose one day after the S&P 500 closed above 6,700 for the first time. A wave of good news from the AI sector lifted chip stocks worldwide, with Nvidia (NVDA) rising to a record high. AMD (AMD) and SK Hynix (000660.KS, HXSCL) also gained.

OpenAI’s (OPAI.PVT) valuation soared to $500 billion after an employee share sale, boosting tech rally hopes despite fears of an AI bubble. The ChatGPT maker ousted Elon Musk’s SpaceX (SPAX.PVT) as the most valuable startup in the world.

Markets have so far been unperturbed by the US government shutdown, which looks set to drag on at least until the end of the week. President Trump is amping up his rhetoric against Democrats, threatening to fire “thousands” of federal workers and canceling billions of dollars in federal funding to blue states.

Trump said he is meeting on Thursday with OMB Director Russ Vought, who has been leading White House strategy in the shutdown, to discuss which “Democrat Agencies” should be cut.

In any case, Friday’s scheduled release of the September jobs report is all but certain to be delayed. That has Wall Street looking elsewhere during the federal data blackout, as Fed policymakers have indicated cracks in the labor market will loom large in their October rate decision.

Private data from the firm Challenger, Gray & Christmas released Thursday found hiring plans at their lowest level since 2009, even as layoffs fell. The report provided more evidence of the softening “low hire, low fire” labor market after Wednesday’s ADP report. Investors remain near-unanimous on bets for a cut at the Fed’s next meeting.

Elsewhere in corporates, Tesla (TSLA) shares took a hit despite a record sales quarter as investors turned focus to future performance without the federal EV tax credit.

Asian shares were mixed on Friday after heavy buying of tech shares led benchmarks on Wall Street to more records.

US. futures and oil prices were higher.

Markets have largely shrugged off the shutdown of the U.S. government after Democrat and Republican lawmakers failed to reach agreement on funding.

U.S. President Donald Trump and congressional leaders were not expected to meet again soon and the Democrats have held fast to their demands to preserve health care funding, warning of price spikes for millions of Americans nationwide.

Japan’s Nikkei 225 rose nearly 1.8% to 45,728.89 as tech stocks gained despite data showing Japan’s unemployment rate rose 2.6% in August, the highest in 13 months and above the expected 2.4%.

Shares in Hitachi jumped 9.2% after it signed a memorandum of understanding with OpenAI to provide cooling systems for its data centers.

Stocks in the computer chip and artificial-intelligence industries also have climbed this week after OpenAI announced partnerships with South Korean companies for Stargate, a $500 billion project aimed at building AI infrastructure.

Stock exchanges in China and South Korea were closed Friday for holidays.

Hong Kong’s Hang Seng index shed nearly 0.9% to 27,052.32, as traders sold to lock in profits from Thursday’s gains.

Australia’s S&P/ASX 200 added more than 0.3% to 8,977.80. India’s BSE Sensex shed 0.2%, while Taiwan’s Taiex rose 1%.

Thursday on Wall Street, the S&P 500 added 0.1% to its all-time high set the day before, closing at 6,715.35. The Dow Jones Industrial Average rose 0.2% to 46,519.72, and the Nasdaq composite climbed 0.4% to 22,844.05.

The government shutdown means this week’s usual report on jobless claims was delayed. An even more consequential report, Friday’s monthly tally of jobs created and destroyed across the economy, will likely also not arrive on schedule.

That increases uncertainty when much on Wall Street is riding on investors’ expectation that the job market is slowing by enough to convince the Federal Reserve to keep cutting interest rates, but not by so much that it leads to a recession.

So far, the U.S. stock market has looked past the delays of such data. Shutdowns of the U.S. government have tended not to hurt the economy or stock market much, and the thinking is that this one could be similar, even if Trump has threatened large-scale firings of federal workers this time around.

That left corporate announcements as the main drivers of trading Thursday.

Excitement around AI and the massive spending underway because of it are a major reason the U.S. stock market has hit record after record, along with hopes for easier interest rates. But AI stocks have become so dominant, and so much money has poured into the industry that worries are rising about a potential bubble that could eventually lead to disappointment for investors.

Still, Advanced Micro Devices climbed 3.5%, and Broadcom gained 1.4%. Nvidia’s 0.9% rise was the strongest single force pushing the S&P 500 upward.

In other dealings early Friday, benchmark U.S. crude added 36 cents to $60.84 per barrel. Brent crude, the international standard, rose 36 cents to $64.47 per barrel.

The U.S. dollar climbed to 147.64 Japanese yen from 147.26 yen. The euro edged up to $1.1725 from $1.1717.

(Reuters) -Universal Music and Warner Music are nearing landmark artificial intelligence licensing deals, the Financial Times reported on Thursday, citing people familiar with the matter.

Universal and Warner could each strike deals with AI companies within weeks, the newspaper said.

Talks involved start-ups such as ElevenLabs, Stability AI, Suno, Udio and Klay Vision, the report said, adding that the music companies are also in talks with large technology groups, including Alphabet’s Google and Spotify.

Reuters could not immediately confirm the report. Universal, Warner, Google and Spotify did not immediately respond to Reuters request for comment.

The growing use of generative AI in creative industries has triggered a wave of lawsuits, with artists, authors and rights holders accusing AI firms of using copyrighted material without consent or compensation to train their models.

The deal talks have centred on how the labels license their songs for creating AI-generated tracks and for training large language models, the report said.

The music companies are seeking a payment structure similar to that for streaming, whereby playing a song triggers a micropayment, the report added.

Taiwan has vowed to resist pressure from Washington to shift half of its chip production capacity to the United States, throwing down the gauntlet to the Trump administration.

Washington has grown increasingly concerned about its heavy dependence on Taiwan, where chip giant TSMC supplies the vast majority of the world’s advanced semiconductors to major clients like AI chip designer Nvidia and Apple.

But many in self-ruled Taiwan view its chip-making prowess a “silicon shield” that could deter a potential invasion from China and rally international support for its security.

Taiwan “will not agree” to producing 50% of its semiconductors – vital to everything from electronics and iPhones to training artificial intelligence and weapon systems – on US soil, Taiwan’s Vice Premier Cheng Li-chiun said Wednesday.

“Our negotiating team has never made any commitment to splitting chips 50-50, so the public can rest assured,” she told media after returning from the US after the latest round of tariff negotiations.

US Secretary of Commerce Howard Lutnick sparked alarm among Taiwanese over the weekend by demanding in a TV interview that Taiwan split its chip production evenly between domestic and US facilities – adding fresh tensions to the ongoing US-Taiwan trade talks. Beyond Taipei’s pushback, opposition party officials and experts have also voiced criticism.

In the NewsNation interview, Lutnick referred to the concept of a “silicon shield,” saying the US would need 50% of domestic chip production to protect Taiwan.

“My argument to them was, well, if you have 95%, how am I gonna get it to protect you? You’re going to put it on a plane? You’re going to put it on a boat?” he said, referring to the rough percentage of global advanced chip production in Taiwan.

“If we have half, we have the capacity to do what we need to do, if we need to do it,” he added, without specifying the response the US may take in the scenario of a Chinese attack.

China claims Taiwan as part of its territory despite having never controlled it and has vowed to annex it, by force if necessary.

Lutnick also said that Washington has been in conversations about the proposal with Taipei, and that the Trump Administration’s objective is to increase the market share of made-in-US chips “to 40% and maybe 50%.”

But Cheng said on Wednesday that the idea did not make it into the latest round of bilateral discussions. It was not clear if TSMC was involved in the negotiations between the US and Taiwan. TSMC declined to comment.

On Monday, Hsu Yu-chen, a legislator of the opposition party Kuomintang (KMT), blasted the US demand as an “outright plunder” instead of “cooperation,” urging the government to reject the demand that she said amounts to “selling out the nation.”

“If the US forces a division of TSMC’s most advanced production capacity, the effectiveness of the ‘Silicon Shield’ will be weakened, and Taiwan’s strategic security leverage will be completely lost,” she said in a statement. “Taiwan needs allies, but not ones who care only about their own security while disregarding Taiwan’s survival.”

On Wednesday, Taiwan’s cabinet said in a statement that the fifth round of trade negotiations just concluded in Washington and led by Cheng has achieved “certain” progress, as officials work toward a goal of reducing US tariffs on Taiwanese goods from the current level of 20%.

Domestically in Taiwan, Washington’s growing list of demands are straining relations with the East Asian economic powerhouse, and risk further souring public sentiment in Taiwan toward the US.

Back in 2020, TSMC responded to US demands and unveiled a groundbreaking $12 billion investment to build advanced chipmaking facilities in Phoenix despite industry and public concerns. Earlier this year, TSMC drastically increased its total investment to $165 billion with additional plants.

These moves have triggered fear and resignation among Taiwanese people, some of whom see the US as using political pressure to rob Taiwan of its proud industry champion and core competitiveness.

Arisa Liu, a director at Taiwan Institute of Economic Research, an independent think tank, said she believes Washington’s latest demand “brings more harm than benefit to Taiwan.”

“Significant investments and capacity shifts toward the US will inevitably weaken Taiwan’s own ecosystem, undermining the integrity of its supply chain,” she said in a statement to CNN, adding that the short-term benefit could be relatively low tariff rates the US imposes on Taiwanese exports.

Taiwan and TSMC owe its success in chipmaking partly to the high concentration of industry players, from silicon wafer suppliers to equipment makers and service providers, experts said. That has contributed to a complete supply chain ecosystem that functions efficiently, a trait required in the manufacturing of semiconductors.