One-Line Summary

Big Tech earnings put an estimated $600B AI capex under scrutiny as platforms roll out new AI ad tools and assistants, and investor money flows into automation plays from data centers to legal tech.

Big Tech

Hyperscaler earnings put AI capex under the microscope

Microsoft, Alphabet, Amazon, and Meta report after the close on Apr 29, and their results are a reality check on the AI buildout: the four “hyperscalers” are expected to spend over $600$ billion this year on data centers and other AI infrastructure, together representing over $10$ trillion in market cap and about 17% of the S&P 500. Options markets are pricing in moves of at least 4%, underscoring how closely investors tie these earnings to the AI trade. ¹

Early investor reactions split: Meta’s stock falls more than 6% after hours, Microsoft trades roughly flat, and Alphabet jumps almost 7%, with Alphabet highlighting “unprecedented” demand for AI compute and raising its 2026 capex guidance to $180$–$190$ billion (from $175$–$185$ billion). The message to the market is clear: spending is climbing, but confidence hinges on proof that AI investments are driving revenue. ²

Cloud performance is the tell: Google Cloud revenue grows 63% year over year to $20$ billion with a $462$ billion backlog that nearly doubles quarter over quarter; AWS posts $37.6$ billion revenue, and Microsoft Cloud reports $54.5$ billion, with Azure and other cloud services up 40%. Microsoft guides that fourth‑quarter capex exceeds $40$ billion and pegs about $25$ billion of 2026 spend to higher component pricing, noting capacity constraints through 2026 as two‑thirds of spend goes to GPUs and CPUs. ²

Another variable hanging over earnings is OpenAI. CNBC reports that AWS adds access to OpenAI’s models, a day after Microsoft and OpenAI announce changes to their partnership; analysts flag both Microsoft’s reliance on OpenAI and OpenAI’s diversification of compute. Private investors value OpenAI at more than $850$ billion, making its momentum a proxy for the broader AI trade. ³

Meta’s business AI tops 10 million weekly conversations

Meta says its business AI tools now handle about 10 million conversations per week across messaging apps as of late March, up from 1 million at the start of 2026—growth the company attributes to a broad beta expansion across the U.S., EMEA, APAC, and LATAM. For small businesses, that means automated replies and customer support are becoming mainstream inside WhatsApp, Messenger, and Instagram. ⁴

The tools remain free for most businesses today, but Mark Zuckerberg signals a path to monetization. On the ads side, Meta reports that more than 8 million advertisers now use at least one generative AI creative tool, and tests show advertisers using its video generation feature see over 3% higher conversion rates. ⁴

Meta also notes its family of apps generates $885$ million in quarterly revenue from paid messaging and subscriptions, while companywide Q1 profit reaches $26.8$ billion on $56.3$ billion in revenue, up 33% year over year—context for why Meta keeps pushing AI into both customer service and ad performance. ⁴

Industry & Biz

SoftBank builds a robotics firm for data centers and eyes a $100B IPO

SoftBank is reportedly forming a new company called Roze AI to make U.S. data center construction more efficient by deploying autonomous robots to help build server farms—an attempt to automate the build‑out behind the AI boom. Some executives want the company prepped for an IPO as early as the second half of 2026 with a potential $100$ billion valuation, according to FT and WSJ reporting cited by TechCrunch. ⁵

The move aligns with a wider push to use AI and automation to speed heavy industry and infrastructure work, similar to efforts like Project Prometheus, which targets modernizing industrial firms with AI. For enterprise teams depending on cloud capacity, this signals greater focus on removing construction bottlenecks. ⁵

Even inside SoftBank there’s reported skepticism about the valuation and the proposed timeline—reminding operators and investors that building physical AI infrastructure at scale is capital‑intensive and execution‑heavy. ⁵

Nvidia and Atlassian back Legora in $600M round at a $5.6B valuation

Nvidia’s venture arm and Atlassian participate in a $50$ million extension to legal AI startup Legora’s Series D, bringing the round to $600$ million and valuing the company at $5.6$ billion. The funding underscores growing strategic interest in AI for complex, document‑heavy workflows. ⁶

Legora says it helps lawyers collaborate with AI on research, review, and drafting, integrating firm data and local legal context. The company reports over $100$ million in annual recurring revenue, claims law firms save an average of 4.3 non‑billable hours per lawyer per week, and notes 42% of customers have won new work using the platform; clients include White & Case, Linklaters, Cleary Gottlieb, and Barclays. ⁷

Atlassian frames the bet as “context‑aware AI” for team collaboration, while Legora’s CEO describes building an “agentic operating system for legal work”—signaling that vendors see autonomous execution with human oversight as the next step beyond assistive chat. ⁸

New Tools

X rebuilds its ad platform with xAI

X begins a phased rollout of a rebuilt, AI‑powered ad platform that replaces its retrieval and ranking systems to deliver more relevant placements and give marketers tighter campaign control. The aim is to make it easier to launch targeted campaigns with better outcomes using xAI’s models. ⁹

The company positions the effort as a full stack rebuild, with X’s ad lead citing a design for rapid, continuous improvements and regular feature drops. Industry coverage describes it as X’s first ground‑up ad infrastructure project since merging with xAI, pointing to deeper model integration over time. ¹⁰

Forecasts suggest X’s ad business is recovering, with eMarketer estimating $2.26$ billion in 2025 ad revenue rising to $2.46$ billion in 2026; the platform’s modernization bid aims to accelerate that trend and compete with AI‑driven ad tools from Google and Meta. ⁹

What This Means for You

For marketers, the bar is moving from “use AI somewhere” to “show lift at scale.” Meta reports 8 million advertisers using its AI creative tools and a >3% conversion uplift for video generation—numbers that justify small A/B tests on your own creative mix to validate impact before broader rollouts. ⁴

If you run paid social, the new X Ads Manager is a low‑risk sandbox: spin up a small, tightly targeted campaign to compare CPA/ROAS against your current baselines and see whether AI‑driven retrieval and ranking improves reach or efficiency. ⁹

For planning and procurement, hyperscaler capex and cloud growth metrics are practical signals. Alphabet’s $20$ billion Google Cloud revenue at 63% growth and a $462$ billion backlog, plus Microsoft’s guidance that capex exceeds $40$ billion in Q4 with capacity constraints through 2026, imply continued tight supply for GPUs and premium compute. Expect lead times and pricing to reflect that. ²

Legal, ops, and IT leaders should evaluate domain‑specific AI where accuracy and confidentiality matter. Legora’s reported $100$ million+ ARR, time savings, and customer logos show legal AI is moving from experiment to workflow—use a pilot on narrow, low‑risk document types (e.g., NDAs) to test fit and controls. ⁷

Action Items

1. Test Meta’s AI video creative on one campaign: Run a controlled A/B against your current best‑performing ad to check if the reported conversion lift appears in your audience.
2. Try X’s rebuilt Ads Manager: Launch a small, audience‑specific campaign to compare CPA/ROAS with your baseline and assess targeting controls and reporting.
3. Pilot a business messaging assistant: If eligible for Meta’s business AI beta, script answers to your top 20 FAQs and route them through WhatsApp or Instagram DMs.
4. Book a 30‑minute legal AI demo: Have your legal or procurement team trial a vendor like Legora or a peer solution on a narrow use case (e.g., NDA review) with scrubbed documents.
5. Skim Alphabet’s Q1 materials for capex signals: Note Google Cloud growth, backlog, and 2026 capex guidance to inform your 12‑month cloud and AI tool budgeting.