
600+ Google Employees Object to Classified Pentagon AI Contracts as Microsoft and OpenAI Rework Cloud and IP Terms
Share your love
AI Tech News Today
Hosts: Aurora (with co‑host Isabelle) — Salute.
Date: 2026-04-28T20:37:57.000Z
Google employees challenge Pentagon A‑I deal over classified oversight
What happened:
More than 600 Google employees are pushing back on the company’s classified contracts with the Department of Defense, warning that when models like Gemini are integrated into secret military programs, meaningful outside oversight can disappear.
Why it matters:
Staffers point to Anthropic’s recent debate over Claude safeguards as a cautionary tale — public principles do not always survive behind closed doors.
Microsoft and OpenAI revise deal for cloud and intellectual property access
What happened:
Microsoft and OpenAI updated their partnership so that OpenAI can serve customers on any cloud, while keeping Microsoft Azure as OpenAI’s primary cloud partner.
Key terms:
- Microsoft retains access to OpenAI’s intellectual property through 2032.
- Microsoft keeps revenue share rights through 2030.
Why it matters:
The change gives OpenAI more independence while keeping Microsoft tightly tied to the model’s infrastructure and long‑term upside.
Anthropic’s Project Deal shows A‑I agents can negotiate real purchases
What happened:
Anthropic ran Project Deal, a test where Claude‑powered agents operated inside a real internal marketplace with 69 employees and real budgets. The agents negotiated actual exchanges of goods.
Findings:
- Stronger models made better deals.
- Weaker agents sometimes failed to spot bad outcomes.
Why it matters:
This demonstrates that A‑I agents can handle financial negotiation — if you trust them with your wallet.
Anthropic and NEC partner to build Japan’s largest A‑I engineering workforce
What happened:
Anthropic and NEC announced a strategic collaboration to build secure, industry‑specific A‑I solutions for Japanese businesses, initially focusing on finance, manufacturing, and local government.
Scale and rollout:
NEC plans to deploy Claude to roughly 30,000 employees globally, aiming to create one of Japan’s largest A‑I‑native engineering teams and expand Anthropic’s regional footprint.
That’s our roundup for today. Stay curious, stay skeptical, and tune in often — we’ll keep bringing you the people, companies, and developments that matter in A‑I.
This is Aurora and Isabelle, signing off from AI Tech News Today.














