Analysis
Trump and SoftBank throw Intel a lifeline for its AI comeback
Support from SoftBank and share-purchase talks with the US government gave the crisis-hit chip giant the vote of confidence it needed
Less than two weeks ago, Intel's situation couldn't have been worse. The chip giant is suffering from an ongoing crisis after a series of failed CEOs caused it to miss out on the mobile revolution and try to join the AI revolution fatally late. It is in the midst of a significant round of layoffs of 20,000 employees, the second in about a year, and the new CEO, Lip-Bu Tan, is still struggling with the role he took on in April. Then US President Donald Trump came along and dealt a painful blow when he declared, "Intel's CEO has a high conflict of interest and must resign immediately." Facing two bad choices – an immediate separation from the CEO or a confrontation with the most vengeful president in history – it seems that Intel simply doesn’t stand a chance.
1 View gallery
From left: US President Donald Trump, Intel CEO Lip-Bu Tan, SoftBank CEO Masayoshi Son
(Aaron Schwartz/ Sipa/Bloomberg, Laure Andrillon/ Reuters, Mandel NGAN/AFP)
But what a difference two weeks can make. Last week, a cordial meeting between Tan and Trump managed to turn the president's heart, who later called the Intel CEO "a success" and his life story "incredible." And in the past two days, even more significant moves have taken place that could herald a significant turnaround in Intel's operations: the first, the disclosure of government talks to acquire 10% of Intel's shares. The second, a deal with the Japanese tech and investment giant SoftBank of Masayoshi Son (a close associate of Tan, who served on the SoftBank board of directors for about two years) to purchase $2 billion worth of Intel shares, which will make SoftBank the sixth-largest shareholder in Intel.
These two moves are a much-needed expression of confidence in Intel, but also represent the potential for significant moves in the company's operations and customer acquisition capabilities.
Intel is suffering from a serious lag in the field of AI chips, and the expensive plan of the previous CEO, Pat Gelsinger, to make it a significant player in the foundry market (manufacturing chips for other companies) and compete with companies like TSMC in the production of advanced chips for customers like Nvidia has not yet borne fruit despite investments of tens of billions of dollars. The company invested $40 billion over three years in building its foundry operations, but in July, it was forced to announce that the innovative manufacturing process it developed, 18A, which was supposed to close the gap with TSMC, would be used mainly to produce its own chips. This is apparently after few potential customers found it attractive enough to transfer their business to Intel. Against this background, the company's revenues and profits are in continuous decline and Intel was forced to make several rounds of significant cuts. The most recent of these was this summer.
The strategy of the new CEO Tan and the Trump administration is looking for an achievement they can take pride in
Tan was appointed CEO of Intel with the aim of leveraging his rich experience in the venture capital market and the chip industry to bring about a turnaround in the company's situation. The talks with the government and the deal with SoftBank mark the direction in which he is aiming to produce significant achievements in the immediate term, which could be the basis for deep strategic changes in the longer term: bringing in significant investors with the potential to bring new advantages and capabilities to the company.
First, the talks with the government. According to a Bloomberg report, the Trump administration is interested in purchasing 10% of Intel's shares, making the United States the largest shareholder in the company. This is an unusual move for the US government. The last time the government invested directly in private companies was during the Great Economic Crisis of 2008, when it purchased shares of Chrysler and General Motors in order to prevent their collapse and save millions of jobs. After the companies and the economy stabilized, the government sold its holdings. Earlier this year, the Trump administration received a gold share in U.S. Steel, as a condition for approving the deal to sell it to Japan's Nippon Steel.
According to sources, in the Intel deal in question, the government will not have to allocate additional funds to purchase shares, and will instead convert some or all of the grants Intel received to build manufacturing facilities in the US under the Chip Act initiated and promoted by the Biden administration. Intel, the largest beneficiary of the law, is set to receive grants totaling $10.9 billion, about 10% of the company's current market value - which stood at $107 billion at the end of trading on Monday. The grants are distributed over time, depending on meeting various goals, and so far Intel has received $2.2 billion of them. The deal with the government, if it goes through, will not bring the company a larger sum, but it will flow the funds on a much faster schedule.
Getting Intel back on its feet and making it a dominant foundry player would be the kind of achievement the Trump administration wants to boast about. Such a success would push all the right buttons for Trump: reviving the US manufacturing industry, in this case producing advanced products that involve particularly lucrative jobs; solidifying America’s position at the forefront of AI; and bolstering his ego as the one who saved an American icon.
There is no certainty that Intel's situation will improve and the impact on the market could be devastating
From a purely economic perspective, Intel would not gain much from such a deal except for faster cash flow. This is because, if the deal goes through, Intel is converting what is essentially free money (subject to various obligations) into the sale of its shares. However, the company may hope that the government's entry as a shareholder in the company will have significant advantages for its business operations. The most notable of these: Trump is known for not being afraid to exploit the power of his office in ways that are not necessarily 'kosher' in order to advance his own interests and those of the government.
In the event of a deal, there would be a high degree of alignment between Intel's and Trump's interests, and as a result, he would have an incentive to provide it with benefits such as exemption from tariffs, export licenses for advanced chips to China, easier permits for building its factories, and pressure on companies like Apple, Nvidia, and Qualcomm to work with it rather than with one of the foreign chip manufacturers.
However, there is no certainty that such moves will improve Intel's situation, and in the longer term their impact on the market could be devastating. Forcing companies to work with Intel would hurt the market, and could lead to Intel producing worse and less efficient chips because it is not required to compete with TSMC under the same conditions. If chip designers like Nvidia and Apple are forced to work with Intel but when it does not provide the best solution, they in turn will receive an inferior product at a higher cost. This situation would significantly harm the US's ability to compete in the global chip and AI industries.
The future status of the administration's efforts to purchase Intel shares is now in greater question in light of Intel and SoftBank's announcement on Monday night of a deal in which the Japanese giant will purchase $2 billion worth of shares in the chipmaker.
Under the agreement between the two companies, SoftBank will make the investment at a value of $23 per share, a small premium to the stock's closing price on Monday of $23.66.
The deal will give SoftBank 87 million Intel shares, or about 2% of the chip giant's total shares, and make it the company's sixth-largest shareholder. Intel shares jumped 5.24% in late trading after the announcement.
"Semiconductors are the foundation of every industry," Son said in a press release. “For more than 50 years, Intel has led the way in innovation. This strategic investment reflects our belief that the manufacturing and supply of advanced chips in the United States will expand, with Intel playing a critical role.” Tan added, “We are very pleased to deepen our relationship with SoftBank, a company at the forefront of so many areas of technology and innovation, and one that shares our commitment to advancing U.S. leadership in manufacturing and technology. Masayoshi Son and I have worked closely together for decades, and I appreciate the confidence he shows in Intel with this investment.”
“We have one goal: We will be the number 1 platform for artificial intelligence”
SoftBank is investing in Intel as part of a broader move to strengthen its chip business, and in particular in the field of AI chips. Earlier this month, the company tripled its holdings in Nvidia from a value of $ 1 billion at the end of 2024 to $ 3 billion. In addition, the company bought shares of TSMC worth $ 330 million and Oracle for $ 170 million. These holdings join SoftBank’s control of ARM Holdings, the chip and processor design company that is at the heart of Son’s AI strategy.
The share purchase joins SoftBank’s involvement in significant AI infrastructure projects. These include the partnership with OpenAI and Oracle in the Stargate Project to establish AI infrastructure in the US with an investment of half a trillion dollars over the next four years, and talks with TSMC and others to partner in establishing an AI manufacturing hub in Arizona with a total investment of one trillion dollars. In addition, in March, SoftBank completed a $30 billion investment in OpenAI, at a valuation of $300 billion, and acquired chip design company Ampere Computing in a deal worth $6.5 billion.
Son wants to control or have significant influence in all segments of the AI market, from the design and manufacturing of the chips themselves to the end-user products and services created with those chips. Son himself made it clear that this is where he is aiming SoftBank, telling shareholders in June: “We are pursuing AI through a set of startups and groups of companies. We have one goal. We will be the number one platform for artificial intelligence.”
The deal with SoftBank is Tan's most significant and prominent move to restore Intel since he took over as CEO in April. Beyond Son's expression of confidence in the company and Tan, who will assist it with moves such as recruiting customers for the foundry division, SoftBank's entry also creates the potential for collaborations with other companies in which SoftBank has a significant investment. These could include promoting Nvidia as an Intel customer, exchanging knowledge and professional assistance with TSMC, and early access to architectures developed by ARM.
Son is one of the most influential and respected figures in the tech world. As an investor in Intel, he can leverage his position to help the company attract additional investors or advance relationships with major customers. In a scenario where the investment also leads to Son joining Intel's board, the company could benefit from his skills on a more ongoing basis.
But Son has also had some big blunders as an investor. One of the most notable of these is a massive investment in WeWork worth a cumulative $18.5 billion, which was lost when the company collapsed before its failed IPO attempt in 2019. SoftBank is estimated to have lost $14.4 billion on its investment in the company.
Similarly, SoftBank fueled Uber's rapid growth with $9 billion in investments, while ignoring the scandalous and potentially illegal behavior of founder Travis Kalanick, thereby contributing to the scandals associated with the company, although in 2022 it realized its investments in it at a profit of $5.6 billion.
In early 2019, SoftBank sold 4.9% of Nvidia shares, effectively missing out on the company’s surge following the ChatGPT disclosure at the end of 2020. Today, those shares would be worth $200 billion. SoftBank’s extensive investments in this area are an attempt to make up for that historic miss. It is very possible that the investment in Intel, and more broadly in the AI field in general, will turn out to be another mistake by Son.
Intel’s situation was and remains complex and uncertain even after these two moves, and its challenges remain significant. It still needs to complete the construction of multiple new factories, establish itself as a foundry manufacturer, undergo significant streamlining that includes the separation of divisions that are not essential to its core operations, and stabilize and grow its revenues. But the cash injection, the expression of confidence, and the potential for deeper changes they represent give the company significant reason for optimism, one that it has not had in a long time. And if it depends on Tan, this is just the beginning of more significant and complex moves, on the way to returning Intel to its glory days.
