Global manufacturing activity strengthened in August, with the artificial intelligence boom driving demand for semiconductors and computing equipment across Asia while a recovery in European orders pushed the euro zone’s factory sector to its strongest level in more than four years.

The improvement offers a more positive picture for global industry, although the outlook remains clouded by the prolonged U.S.-Iran conflict, disruptions around the Strait of Hormuz and rising input costs.

The euro zone recorded one of the strongest improvements. S&P Global’s Eurozone Manufacturing Purchasing Managers’ Index rose to 52.7 in August from 51.9 in July, marking its highest reading since May 2022. The figure was slightly below the preliminary estimate of 52.8 but remained comfortably above the 50 mark that separates expansion from contraction.

Germany led the recovery, with its manufacturing PMI climbing to 54.3, its strongest reading in more than four years. France also returned to expansion, with its PMI rising to 51.1 from below 50 in July.

The improvement was not uniform across the region. Italy recorded its first manufacturing contraction since January, while Spain also remained in contraction territory, suggesting that the euro zone’s recovery is still uneven.

“The resilience story is still going on,” said Carsten Brzeski at ING.

He said some European manufacturers were benefiting from the disruption affecting Asian competitors as the closure of the Strait of Hormuz altered global trade flows.

“It still reflects the fact European manufacturing companies, at least some of them, benefited from the fact Asian competitors are hurt more by the closure of the Strait of Hormuz,” Brzeski said.

The shipping disruption has weighed heavily on manufacturers because the Strait of Hormuz carried about a fifth of global oil supplies before the war began in late February. Efforts by Qatar and Oman to broker an agreement that would reopen the waterway have so far failed to produce a breakthrough.

For European factories, however, the stronger August readings come against a difficult energy backdrop. Higher oil and gas prices raise production and transportation costs, creating a risk that the manufacturing recovery could lose momentum if the conflict continues to disrupt energy markets.

Britain also remained in expansion, although its manufacturing PMI slipped to 51.7. A more encouraging feature of the British survey was employment, with factories increasing hiring at the fastest pace in more than two years as production requirements increased.

AI Drives Asian Manufacturing

Asia presented a stronger picture, with China, Japan and South Korea all recording manufacturing growth as demand for AI infrastructure, semiconductors, computers and related equipment supported industrial production.

China’s RatingDog China General Manufacturing PMI, compiled by S&P Global, rose to 51.5 in August from 50.9 in July and exceeded the 51.0 median forecast in a Reuters poll.

The improvement points to the growing importance of AI-related demand in supporting China’s industrial economy. The picture remains fragile, however, as a separate official survey showed China’s broader factory activity remained in contraction.

Japan recorded an even stronger acceleration. Its manufacturing PMI rose to 54.9 from 54.5, the highest level since April, while new business expanded at its fastest pace since January 2018. The strength of new orders significantly indicates that the Japanese manufacturing recovery is being supported by actual demand rather than simply inventory rebuilding.

“Overall, the sector looks well placed to sustain its strong performance, particularly given demand linked to AI-related sectors,” said Annabel Fiddes, economics associate director at S&P Global Market Intelligence.

South Korea’s manufacturing PMI eased to 52.3 from 53.1 but remained above 50 for a ninth consecutive month. Separate trade data provided further evidence of the strength of the country’s technology exports.

South Korean exports surged 68.7% from a year earlier in August, extending their growth streak to 15 consecutive months. Semiconductor and AI-related products have been among the principal beneficiaries of the global technology investment cycle.

The contrasting regional performances highlight an increasingly important feature of the global manufacturing recovery: AI investment is creating a concentrated source of industrial demand even as conventional manufacturing remains exposed to trade tensions, energy costs and geopolitical disruptions.

For chipmakers and electronics manufacturers in Asia, the AI investment cycle is generating orders for advanced processors, memory, servers and networking equipment. That demand is helping offset weakness elsewhere in the industrial economy. For Europe, the recovery is coming from a different direction, with improving new orders and reduced competitive pressure providing some relief to manufacturers.

However, analysts believe the sustainability of the global upturn hangs on how strong AI demand remains to offset the drag from higher energy costs and geopolitical uncertainty. A prolonged disruption around the Strait of Hormuz could eventually feed into production costs worldwide, while any slowdown in technology investment would expose manufacturers that have become dependent on the AI spending cycle.

South Korea has unveiled its largest-ever fiscal spending increase, proposing an 821 trillion won ($596.9 billion) budget for 2027 as President Lee Jae-myung’s government seeks to use the country’s semiconductor windfall to accelerate investment in artificial intelligence, advanced manufacturing and strategic technologies.

The budget proposal represents a 12.8% increase from this year’s spending and marks the biggest year-on-year expansion on record, according to the budget ministry. The scale of the increase signals a decisive shift toward fiscal expansion after three years of austerity under Lee’s predecessor.

The government is betting that exceptionally strong profits from Samsung Electronics and SK Hynix, driven by surging global demand for high-bandwidth memory chips used in AI systems, will provide enough additional tax revenue to finance higher spending while simultaneously reducing new borrowing.

Total tax revenue is projected to rise 40.7% to 584.4 trillion won in 2027. Corporate tax receipts are expected to more than double to 216.7 trillion won, reflecting the sharp improvement in earnings among South Korean companies, particularly semiconductor manufacturers.

The stronger revenue outlook is projected to lower the government’s debt-to-GDP ratio to 48.3% next year from an estimated 51.6% in 2026, a reduction of 3.3 percentage points.

The fiscal strategy therefore differs from a conventional stimulus financed primarily through additional debt. Seoul plans to increase overall spending substantially while using higher tax receipts to restrain the growth of government borrowing.

Semiconductor Windfall Drives Investment Push

A major focus of the budget is strengthening South Korea’s position in the global AI supply chain.

The government has allocated 21.3 trillion won for infrastructure supporting next-generation semiconductor manufacturing, including industrial water systems, electricity grids and logistics networks. It has also earmarked 2.6 trillion won for a dedicated semiconductor budget, according to the cabinet presentation.

The investment comes as South Korea’s chipmakers benefit from one of the strongest periods of AI-related demand in the semiconductor industry. Samsung and SK Hynix are major suppliers of HBM, a critical component for the high-performance processors used to train and operate advanced AI models.

The government is effectively seeking to convert that cyclical export and corporate-profit boom into longer-term industrial capacity. Reliable electricity, water and logistics infrastructure have become crucial as chipmakers build larger and more sophisticated fabrication and packaging facilities.

But the strategy is also seen as a reflection of the intensifying competition among major economies to secure domestic control over critical parts of the AI supply chain, from advanced memory and processors to data-center infrastructure.

Future Response Fund Targets Long-Term Growth

Rather than spend the projected 162.3 trillion won tax windfall entirely through conventional government programmes, Seoul plans to channel it into a strategic endowment known as the Future Response Fund.

The fund is intended to finance longer-term priorities, with 45.4 trillion won scheduled for deployment in 2027 toward youth welfare, future growth industries and specialized education. That approach gives the government a mechanism to convert unusually strong semiconductor tax receipts into investments intended to support future productivity rather than simply expanding recurring expenditure.

The challenge will be ensuring that the spending generates sufficient economic returns. South Korea is highly exposed to global technology cycles, and semiconductor earnings can fluctuate sharply when demand, prices or inventories change.

However, the budget extends beyond technology and industrial infrastructure.

The government proposed 3.4 trillion won for a nuclear-powered submarine programme and other strategic weapons, according to presentation materials from the cabinet meeting. The increased defense allocation comes as Seoul seeks to strengthen its strategic capabilities while maintaining its position as a major technology and manufacturing power in Asia.

The combination of semiconductor investment, AI-related infrastructure and defense spending indicates that the government is treating technological capacity as an economic as well as a national-security priority.

Bond Market Remains Cautious

Despite the government’s effort to limit new borrowing, South Korean bond yields rose following the budget announcement. The 10-year government bond yield increased 6.5 basis points to 4.378%, suggesting investors had expected a larger reduction in government bond issuance.

Total government bond sales are planned at 222.8 trillion won in 2027, slightly below the 225.7 trillion won planned for this year. Net bond issuance will fall more substantially, to 96.3 trillion won from 109.4 trillion won.

“It would have been better for the market if the government made a bigger reduction (of bond sales),” said Kong Dong-rak, an analyst at Daishin Securities.

Kong said the decline in net issuance was positive but added that adjustments to reduce the allocation of longer-dated debt would help stabilize the domestic bond market.

The reaction highlights the tension facing Seoul. The government is simultaneously pursuing aggressive fiscal investment and trying to contain pressure on the sovereign bond market at a time when global long-term yields are rising.

The budget also comes as South Korea confronts a more difficult monetary-policy environment. President Lee said Tuesday that the economy had reached a point where an interest-rate increase was unavoidable, a move that could weigh on growth by raising borrowing costs for households and businesses. The development has created a potential conflict between fiscal and monetary policy. The government wants to increase investment and support future growth, while higher interest rates would tighten financial conditions to address inflation and financial stability concerns.

For Seoul, the semiconductor boom provides an unusually favorable opportunity to finance that investment without significantly increasing the debt burden. But the strategy also carries risks if chip-sector profits weaken before the infrastructure and technology investments begin generating returns.

The 2027 budget is therefore more than a record spending plan. Economists say it represents Lee’s attempt to use a semiconductor-driven revenue windfall to reposition South Korea for the next phase of the global technology race, while keeping debt and borrowing under control.

The proposal must still be approved by parliament before becoming law.

OpenAI has rejected Apple’s allegations that it stole the iPhone maker’s trade secrets,, arguing that Apple has failed to identify confidential information that was misappropriated by former employees who joined the ChatGPT developer.

In a filing late Monday in U.S. District Court in San Jose, California, OpenAI described Apple’s lawsuit as the product of the company’s own internal procedures and accused it of trying to discourage employees from joining a growing competitor.

“This dispute is a mess of Apple’s own making, and it is trying to blame everyone else,” OpenAI said in the filing.

Apple sued OpenAI and former Apple employees Tang Tan and Chang Liu in July, alleging that they misappropriated trade secrets involving hardware design, manufacturing and supply-chain operations as OpenAI expands into consumer devices.

The case has opened a new front in the rivalry between the two technology companies, whose relationship has deteriorated sharply since they announced a partnership two years ago. Apple had integrated ChatGPT into parts of its ecosystem as it sought to strengthen its artificial intelligence capabilities, while OpenAI has increasingly moved toward developing its own consumer hardware.

OpenAI said Apple is now attempting to use litigation to slow that expansion and discourage its employees from leaving for a potential competitor. The companies have disclosed in court filings that OpenAI has hired roughly 400 former Apple employees for its hardware initiative.

At the center of the dispute is whether former Apple employees improperly retained or used confidential information after joining OpenAI.

Apple has alleged that Tan and Liu accessed internal company files after leaving Apple. The company has also presented new evidence concerning Liu, alleging that he accessed a power-converter circuit schematic while working at OpenAI and used proprietary Apple information to train an AI agent in March 2026.

Apple said the evidence emerged from a MacBook that OpenAI provided to the company on August 21 as part of the litigation. Apple is seeking to accelerate discovery, which would allow the parties to obtain and examine additional documents and other evidence.

OpenAI, however, disputes Apple’s characterization of the conduct.

The company said California law permits employees to move between competing companies and argued that Apple’s own procedures made it difficult for departing workers to separate personal information from company documents clearly.

According to OpenAI, Apple encourages employees to use personal iCloud accounts to access work documents and perform their duties. It also argued that Apple’s practice of immediately escorting departing employees from its premises can leave little time for them to return devices, transfer files, or complete the handover of responsibilities.

Liu said in the filing that any access to Apple documents after his departure was intended to help former colleagues locate files or answer questions relating to Apple’s work. He said Apple employees continued contacting him for assistance after he left.

Tan, who worked at Apple for 24 years, said he returned Apple prototypes before departing and retained only materials that were not confidential, including an employee departure checklist.

OpenAI said Apple’s objections to employee departures do not establish that trade secrets were stolen.

“Employees can leave a company like Apple that has struggled to adopt AI and move to an exciting startup that builds innovative products,” OpenAI wrote. “Apple may not like those choices. But it cannot claim those choices are unlawful.”

The dispute has drawn significant attention because OpenAI’s hardware ambitions could eventually place it in more direct competition with Apple. OpenAI has been assembling a substantial hardware team and pursuing consumer devices designed around its AI systems, potentially challenging Apple’s position at the intersection of hardware, software and consumer technology.

The hiring of hundreds of Apple employees has also given the dispute a broader competitive dimension. OpenAI recruiting engineers and executives with experience in Apple’s highly integrated hardware operation can accelerate its effort to build devices. For Apple, the departure of personnel with knowledge of its hardware development processes raises concerns over the protection of proprietary information.

Apple has sought to frame the issue as more than ordinary employee mobility. Its latest allegations focus on whether confidential technical information was accessed or used after an employee moved to OpenAI.

OpenAI, meanwhile, has sought to distinguish lawful recruitment from unlawful use of trade secrets and has argued that Apple’s internal document-management practices complicate its claims.

The court will ultimately have to determine whether Apple can substantiate its allegations and establish that protected information was improperly acquired, retained or used. OpenAI has separately asked the court to dismiss Apple’s lawsuit, arguing that the products it is developing are “entirely new.”