Amazon Trims AGI Unit While Committing $200B to AI Infrastructure in 2026

Amazon is reorganizing its AI ambitions, cutting roles in its artificial general intelligence unit even as it prepares to spend $200 billion on AI infrastructure this year — a clear signal that scale of investment and size of team are no longer moving in the same direction.

The company confirmed job cuts in its AGI organization on Wednesday but did not disclose the number of employees affected. The AGI unit handles frontier AI models and teams working on custom AI silicon (Trainium and Inferentia chips) and quantum computing.

Key Takeaways

• Amazon's AGI layoffs come alongside $200B in projected 2026 capital expenditure — a 50%+ increase from 2025 — focused on AI data centers, chips, and cloud infrastructure.
• The cuts reflect a reallocation toward commercial priorities: Peter DeSantis now leads the AGI org, signaling closer integration with AWS cloud strategy.
• Amazon acknowledges its models "haven't been at the very frontier" for the most demanding workloads — a candid admission driving the current refocus.
• Through AWS, Amazon offers enterprises access to both its Nova model family and Anthropic's Claude, keeping workloads in its cloud ecosystem regardless of model choice.

An Amazon spokesperson framed the cuts as strategic sharpening: "We're sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts."

This reflects the broader pattern across big tech: massive AI capital expenditure paired with targeted workforce reductions, as companies distinguish between exploration-phase AGI research and commercially viable AI products.

Read the full article on Tekedia