Earnings week puts AI capex and model rollouts under scrutiny at Microsoft and Alphabet
Enzo Reno ·
Microsoft and Alphabet will open a pivotal earnings stretch this week in the United States, with artificial intelligence infrastructure spend and model deployment progress set to dominate investor and customer questions. Their guidance will shape vendor negotiations and 2027 budget plans as enterprises weigh where to build and buy AI capabilities.
The stakes are clear in prior disclosures. Microsoft reported 14.4 billion dollars in capital expenditures in the June 2024 quarter and signaled higher spend to support AI, according to CFO Amy Hood. Alphabet posted 13.2 billion dollars of capex in the same quarter and said 2024 would be an elevated investment year, per CFO Ruth Porat. Meta lifted its 2024 capex outlook to 35 to 40 billion dollars to fund AI, the company said on its second quarter call.
H2: What to watch in guidance
Executives will be pressed on compute availability and cost curves. NVIDIA began shipping its Blackwell generation in late 2024, and data center revenue surged past 26 billion dollars in the July 2024 quarter, the company reported. AMD’s MI300 family started ramping across cloud and OEMs in 2024, and hyperscalers are expanding custom silicon stacks, including Microsoft’s Maia and Cobalt, Google’s TPU v5p, and AWS Trainium and Inferentia, as previously announced.
Enterprises also want clearer signals on AI product uptake and margins. Watch for updates on Microsoft Copilot attach rates across Office and GitHub, Gemini usage in Google Workspace and Cloud, and Meta’s AI features engagement, along with pricing or inference efficiency changes that can shift gross margin. Cloud market share context remains stable, with AWS at about 31 percent, Microsoft Azure at 25 percent, and Google Cloud at 11 percent of global cloud infrastructure services as of mid 2024, according to Synergy Research Group.
H2: Policy calendar to monitor
Compliance pressure in Europe is rising. The EU AI Act entered into force in 2024, with rules for general purpose AI applying after 12 months and core obligations for high risk systems beginning after 24 months, according to the European Commission. That means many providers are now in the enforcement window for risk management, data governance, and transparency, which will inform enterprise procurement checklists and contract terms this quarter.
For technology leaders, this week’s earnings calls are not about headline models alone. They indicate where the AI stack will consolidate, how fast inference costs can fall, and which clouds can deliver capacity and compliance at scale, outcomes that will drive startup roadmaps and large enterprise deployment choices through 2027.