'AI Over Earnings'... Meta Platforms Bets on 210 Trillion Won Infrastructure Investment [Global AI Briefing]

[Financial News] Although concerns about profitability are spreading on Wall Street as artificial intelligence (AI) companies ramp up investment, AI firms appear to be betting on expanding AI infrastructure rather than near-term earnings.
On the 29th local time, Meta Platforms announced its second-quarter 2026 results and raised its capital expenditures (CAPEX) outlook for this year from $125 billion to $145 billion, or about 181 trillion won to 210 trillion won, to $130 billion to $145 billion, or about 188 trillion won to 210 trillion won.
The key point was that while the upper end remained unchanged, the minimum investment amount was raised by $5 billion. Market watchers interpreted this as a signal that Meta Platforms intends to spend at least $130 billion under any circumstances.
After the earnings release that day, Meta Platforms' share price plunged more than 10%.
By formally committing to expand AI infrastructure investment despite Wall Street's concerns over profitability, Meta Platforms effectively reaffirmed its strategy of prioritizing long-term AI competitiveness over short-term performance.

Meta Platforms said most of the investment will go into AI data centers, GPU servers, next-generation networks, massive AI clusters, and inference infrastructure. As generative AI services increasingly require enormous computing resources at the inference stage for actual service delivery rather than training, data centers and AI servers are emerging as core assets that determine corporate competitiveness.
After the earnings release, Meta Platforms Chief Executive Officer (CEO) Mark Zuckerberg made clear at an investor briefing that "the next AI competition will be won not only by the company with the best model, but also by the company with the most powerful computing infrastructure." He also outlined a vision of developing personal AI assistants, AI agents, and eventually AI computing services as new growth engines for Meta Platforms.
Recently, major U.S. tech companies have been signaling that they will expand AI infrastructure investment despite Wall Street's concerns.
On the 22nd, Google parent Alphabet Inc. raised its annual CAPEX guidance for this year from $180 billion to $190 billion, or about 260 trillion won to 275 trillion won, to $195 billion to $205 billion, or about 282 trillion won to 296 trillion won, effectively declaring an additional investment of about $15 billion, or about 22 trillion won, compared with the plan it had presented just three months earlier.
Alphabet Inc. also said, "Demand for AI infrastructure is exceeding supply capacity, so we decided to accelerate the expansion of data centers and servers," adding that investment would increase meaningfully again in 2027. That day, Alphabet Inc.'s share price also plunged, underscoring Wall Street's concerns about profitability.
In the end, big tech companies such as Alphabet Inc. and Meta Platforms appear to view how quickly they can secure massive AI infrastructure as the new battleground in the AI era, and are pouring huge amounts of investment into that race.
[email protected] Lee Gu-soon Reporter