One-Line Summary

Big Tech keeps raising AI infrastructure budgets while companies restructure around AI and the industry splits over open-weight models.

Big Tech

Big Tech expands AI buildout amid investor jitters

The New York Times reports that Amazon, Google and other major tech firms are increasing spending on AI data centers and chips even as concerns about payback grow. For everyday users, this means faster rollouts of AI in the tools you already use, powered by more compute behind the scenes. ¹

For business teams, accelerated AI investment often comes with tougher questions from finance about return on investment: which features drive revenue, cut support tickets, or speed content creation. Expect more pressure to show measurable outcomes tied to AI usage rather than experiments without clear metrics. ¹

Watch upcoming capex disclosures and cloud pricing for signals on how aggressively this buildout continues and whether management emphasis shifts toward profitability over usage growth. Those signals will shape how much freedom product and marketing teams have to experiment with AI at scale. ¹

Industry & Biz

Monday.com cuts 20% as firms cite AI in restructuring

TechCrunch reports Monday.com will lay off about 20% of its workforce—just over 600 roles—as it reorganizes around an “AI-driven growth strategy” across product, marketing, and go-to-market; the company expects $45 million to $55 million in restructuring charges and still projects up to 20% year-over-year revenue growth for 2026. For teams, that points to AI skills becoming a hiring filter even when headcount falls. ²

According to TechCrunch’s roundup (citing Financial Times analysis), U.S. tech companies have cut nearly 140,000 jobs in 2026 as of Jul 25, with Amazon, Oracle, Meta, and Microsoft accounting for almost 50,000; companies that cite AI as a factor reportedly underperform the Nasdaq by almost 10% in the 30 trading days after their announcements, suggesting investors want clearer ROI proof. ²

Nvidia leads new alliance as industry backs open-weight AI

Axios reports that Nvidia, Microsoft, Meta, Palantir and dozens of others signed a letter supporting “open-weight” AI models—systems where model weights are released so organizations can run them internally—with Google and OpenAI also joining; Anthropic has not signed. This debate shapes whether enterprises can self-host models for control and auditing versus buying access to closed APIs. ³

Axios adds that Nvidia launched the Open Secure AI Alliance to promote open research and security for such models while noting “the world needs both closed and open models,” amid U.S. policy discussion over rapidly improving Chinese open-weight models. For buyers, the choice affects cost, security review, and vendor dependence. ³

What This Means for You

As AI capex rises, CFOs will ask for proof that AI features move core metrics like conversion, retention, ticket deflection, or time-to-ship. Teams that tie AI usage to quantifiable outcomes will defend budgets better than those pitching generic productivity gains. ¹

Restructuring trends mean some roles shrink while AI-adjacent work expands. Document how AI changes your workflow (e.g., drafting, QA, reporting) and show measurable cycle-time cuts or quality lifts—evidence that strengthens your role even as organizations streamline. ²

The open-weight push signals more options for running models where your data lives, but it also raises governance questions. Coordinate with IT/security on data control, auditability, and incident response when comparing closed APIs versus open-weight deployments. ³

Action Items

1. Publish a one-page AI ROI brief: List the top three AI-enabled features you ship or use, the target metric for each (conversion, CSAT, time saved), and how you’ll measure it this month.
2. Run a 7-day AI usage audit: Export invoices or usage logs from your AI tools and estimate cost-per-task or cost-per-user; flag two workflows where cost is high and results are weak.
3. Request an IT-reviewed open-weight trial: Ask security/IT to scope a small, compliant pilot of an open-weight model (private environment, limited data) and compare data control and latency with a closed API.
4. Redesign one workflow for AI assist: Pick a recurring task (e.g., campaign brief, QA checklist), define the AI-assisted steps, and set success criteria (minutes saved, errors reduced).
5. Brief leadership on model sourcing: Create a one-slide comparison of closed API vs open-weight options covering data handling, auditability, cost predictability, and support needs.