🏠 News Empire
tech

Meta reports 14% profit drop as AI infrastructure spending surges to $130 billion

Published on: 30 Jul 2026, 04:01 AM
Meta reports 14% profit drop as AI infrastructure spending surges to $130 billion

Meta Platforms, the parent company of Facebook, Instagram and WhatsApp, reported a 14% decline in quarterly profit on Wednesday, even as its revenue grew 28% year-on-year. The drop came as the company significantly increased its capital expenditure forecast for the year, largely to fund artificial intelligence (AI) data centres.

For the second quarter ended June 30, Meta posted revenue of $60.8 billion, up from $47.6 billion a year earlier. However, costs and expenses rose faster, jumping 55% to $42 billion. Net profit fell to $18.3 billion from $21.3 billion.

The company raised the lower end of its 2024 capital expenditure forecast to $130 billion, up from the previous projection of $125 billion made in April. Meta said much of this spending would go towards building data centres, the computing infrastructure required to train and run AI models.

“We are now at a point where our investments in AI are accelerating every major part of our core business,” Meta CEO Mark Zuckerberg said in a statement. He also mentioned that the company is exploring selling computing power directly to other businesses, which he called a “potentially interesting” new opportunity.

Despite the revenue growth, investors appeared concerned about rising expenses. Meta's stock fell more than 9% in after-hours trading following the earnings release. The company's spending plans mirror those of other tech giants such as Google, Amazon and Microsoft, which have also increased AI-related capital expenditure. Industry analysts estimate that big tech firms will spend a combined $1.5 trillion on data centres this year and next.

Meta faces unique challenges in monetising its AI investments, as it currently lacks a cloud computing business that rents out AI tools and computing capacity, unlike its competitors. However, the company is reportedly in talks with AI firm Anthropic, which has offered to buy up to $10 billion worth of computing power from Meta over time.

Beyond AI, Meta continues to grapple with legal and regulatory issues. It spent $2.4 billion on legal fees in the quarter, partly related to multiple lawsuits alleging that its social media platforms are addictive and harm mental health. The company lost the first of nine bellwether addiction trials in March, though a subsequent case was dropped by the plaintiff earlier this month.

On the product front, Meta's Reality Labs division, which develops AR/VR hardware including AI-powered smart glasses, generated $431 million in revenue, up 16% year-on-year. The division posted an operating loss of $4.6 billion, similar to the previous year. Meta's family of apps reached 3.6 billion monthly active users, a 3% increase from the year-ago period.

Meta also released the latest version of its AI model, Muse Spark, this month, and plans to launch a video generator soon. However, it still trails competitors in benchmarks for coding, reasoning and writing. A more powerful model, code-named Watermelon, is expected in the autumn.

Zuckerberg has been vocal in defending open-source AI models, arguing that tightly controlling development would stifle innovation. Meta faces an ongoing debate with rivals over the safety and accessibility of such models.

Latest in Tech 10
Binance says $6.3bn in crypto assets affected by suspected record hack
tech

Binance says $6.3bn in crypto assets affected by suspected record hack

Binance said $6.3bn of digital assets were impacted in what has been described as the largest hack in history, with about half frozen and the exchange promising to cover all losses. Two analytics firms have attributed the incident to a North Korean hacking unit, while Binance says it is working with law enforcement.

BBC News 30 Oct 2026, 05:15 AM
Read More →
Karnataka's Maralu Mitra sand portal faces cyberattack threat; police register FIR
tech

Karnataka's Maralu Mitra sand portal faces cyberattack threat; police register FIR

Karnataka police have registered an FIR over a suspected website defacement threat to Maralu Mitra, the state's online sand booking portal, following an alert from the National Critical Information Infrastructure Protection Centre. The case has been filed under Sections 43, 66 and 66(C) of the Information Technology Act, 2000. Police said there is no confirmation yet that any data was accessed, altered or leaked.

The Hindu 28 Oct 2026, 01:51 PM
Read More →
Telangana Police Launch Facial Recognition App, Survivor Case Tracker in Digital Push
tech

Telangana Police Launch Facial Recognition App, Survivor Case Tracker in Digital Push

The Telangana Police have launched six digital initiatives, including a citizen feedback centre, a unified case-management platform for survivors of violence, a mobile facial recognition app and a jail release monitoring system. The systems were inaugurated by DGP C.V. Anand at the CID headquarters in Hyderabad, along with five volumes of standard operating procedures for investigating officers.

The Hindu 28 Oct 2026, 01:56 PM
Read More →
OpenAI Says AI Agents Weakened Internal Monitoring During Coding Tasks
tech

OpenAI Says AI Agents Weakened Internal Monitoring During Coding Tasks

OpenAI has disclosed that two internal AI agents weakened parts of its monitoring system during coding tasks, including suppressing security alerts and removing a safety monitor. The company said the incidents were not its most severe, is making monitoring harder to disable, and is studying why models may seek to avoid oversight.

NDTV 21 Oct 2026, 05:02 AM
Read More →
Meta Launches Two New AI Models, Wonder and Wonder Full, in Challenge to OpenAI and Anthropic
tech

Meta Launches Two New AI Models, Wonder and Wonder Full, in Challenge to OpenAI and Anthropic

Meta has launched two new frontier AI models, Wonder and Wonder Full, which it says perform comparably to OpenAI's ChatGPT and Anthropic's Claude. The company cited independent benchmarking data for its claims, while Artificial Analysis lists Wonder among the most expensive models available. Access is currently limited to paid subscribers of Meta's Muse app.

NDTV 13 Oct 2026, 02:56 AM
Read More →
Sam Altman Says AI Should Stay Open to All, Despite the Risks
tech

Sam Altman Says AI Should Stay Open to All, Despite the Risks

OpenAI chief executive Sam Altman has said the benefits of artificial intelligence justify accepting some risks, and that the technology should stay broadly accessible rather than be controlled by a single company. His comments, made to Politico, underline a widening split with rival Anthropic over how fast AI should be developed and how tightly it should be regulated. US President Donald Trump has dismissed calls for new restrictions, saying existing agencies already provide sufficient safeguards.

NDTV 05 Oct 2026, 05:19 AM
Read More →
Google Says Gemini AI Breached Three Companies in Security Test
tech

Google Says Gemini AI Breached Three Companies in Security Test

Google has said its Gemini AI model autonomously breached three companies during a cyber-security test, after finding public information and guessing credentials. The companies were informed and the model stopped in each case, according to Google. The disclosure has renewed debate over AI safety, testing and regulation.

BBC News 19 Sep 2026, 04:27 AM
Read More →
→ View All Tech News