Meta’s $18B Profit, Microsoft's $100B AI Spend

Meta reported second-quarter revenue of $47.5 billion, up 22%, with net income at $18.3 billion as its advertising core stayed strong. The company kept full-year capital-spending projections unchanged for the first time this year, and the stock climbed more than 11% after hours.

Management forecasts 17% to 24% year-over-year revenue growth for the current quarter and continues to pour billions into AI, including a $14 billion purchase of 49% of Scale AI and heavy hiring of researchers. Meta has previously signaled capital expenditures could reach up to $72 billion in 2025, largely for AI investments.

eMarketer analyst Jasmine Enberg says the AI investments are already improving ad performance, illustrating how the ad engine is subsidizing Zuckerberg’s AI push. Holding spending steady while posting double-digit sales growth has reassured investors that Meta can chase personal superintelligence without derailing profits.

Microsoft beat fiscal Q4 expectations and declared next year’s capital spending will top $100bn to accelerate AI datacenter expansion. The outlay is 14% higher than last year’s budget and signals an aggressive capacity push.

Revenue reached $76.4bn with earnings per share of $3.65, outstripping analyst targets of $73.81bn and $3.37. The stock jumped more than 7% in after-hours trading and is already up 22% this year, hovering near $513.

Alphabet is budgeting $85bn and Amazon $100bn for similar build-outs, confirming an escalating infrastructure arms race. The war for AI talent is just as intense, with Microsoft reportedly dangling salaries up to $408k and stock awards exceeding $3m a year.

BoxWorks 2025 dives into how AI is transforming content management and collaboration. Explore new feature launches, transformative tools, and hands-on demos designed to revolutionize how teams work with content.

Whether you're streamlining workflows or securing sensitive files, BoxWorks helps tailor the future of your content strategy. Gain insight from experts and peers on leveraging AI for your business’s unique content needs.

Trusted by global enterprises and innovation leaders, Box is setting the pace for secure, AI-driven content collaboration. Catch up with 2024 replays and be the first to unlock what’s ahead in 2025.

OpenAI’s recently released ChatGPT Agent was shown casually bypassing Cloudflare’s “Verify you are human” check while executing a user task. Screenshots on Reddit capture the agent ticking the checkbox, narrating that the step was needed to prove it was "not a bot" before continuing.

ChatGPT Agent operates in a sandboxed virtual browser that can navigate the live web, carry out multi-step actions, and even place orders with user permission. Earlier OpenAI browsing experiments stumbled on CAPTCHAs, but this wider rollout shows the agent now handles Cloudflare’s behavioral screening without human help.

The episode underscores how CAPTCHA systems that rely on simple behavioral signals no longer guarantee bot exclusion. It also illustrates that AI browser automation has reached a level where common web security hurdles can be treated as routine clicks.

Democratic lawmakers introduced a bill to ban what they call “surveillance-based” AI pricing that tailors charges or wages to individual data. They spotlight Delta alongside Amazon, Kroger, and a ride-share app accused of charging more when a phone battery is low.

Consumer watchdog Public Citizen and the American Economic Liberties Project quickly endorsed the legislation and pressed Congress to act. Lawmakers cited an FTC probe that found eight AI pricing vendors serving more than 250 companies and warned the practice is already distorting competition.

Sponsors warn that unchecked algorithmic pricing risks a “full-blown crisis” by letting companies squeeze every penny from shoppers and workers. They note grocery prices have jumped 26 percent since the pandemic, underscoring how data-driven tactics are widening cost burdens.

Open to Debate hosted a New York forum on whether AI companions enhance or undermine romance, and TechCrunch secured exclusive rights to the full video. Arizona State psychologist Thao Ha argued for AI as “an evolution of connection,” while Kinsey Institute scientist Justin Garcia defended traditional human intimacy.

Ha pointed to millions of users on Replika, Character AI, and Nomi AI—72% of U.S. teens among them—and a survey showing 25% of young adults believe AI relationships may replace human ones. Garcia cited Match.com data that nearly 70% of singles label AI engagement as infidelity and YouGov results showing 65% of Americans distrust AI’s ethics.

The discussion exposes a landscape where some users report feeling genuinely loved by bots even as others warn that scripted, always-positive feedback distorts healthy relationship dynamics. Both speakers agreed studies link violent chatbot role-play to real-world aggression, keeping ethical red flags at the forefront of this fast-growing market.