Key Takeaways
- Amazon is in advanced talks to acquire AI startup Decart for roughly $7 billion, a deal that would top its $6.5 billion MGM purchase.
- Decart’s software lets developers shift workloads across different AI chips, and it is building world‑modeling tools (Lucy and Oasis) that generate simulated 3‑D environments.
- The acquisition would give Amazon early access to its own Trainium3 chip, positioned to challenge Nvidia’s dominance in AI‑focused GPUs.
- Prior reports indicate Anthropic walked away from a $6 billion Decart deal, and Israeli outlet Calcalist said Decart was in advanced talks with a major tech firm.
- Amazon is simultaneously expanding its Seller Assistant for third‑party merchants, evolving it from a simple chat tool into an operating layer that can reason across inventory, pricing, advertising, demand and compliance.
- Mary Beth Westmoreland, VP of Worldwide Selling Partner Experience, said the assistant can “remember how a merchant operates; continuously monitor the business; and take action when granted permission.”
- If completed, the Decart purchase would deepen Amazon’s AI infrastructure while its Seller Assistant aims to reduce the frequency with which sellers need to ask questions, turning AI into core business infrastructure.
Amazon’s Potential $7 Billion Acquisition of AI Startup Decart
According to a Sunday Oct 11 report by The Wall Street Journal (WSJ), citing sources familiar with the matter, Amazon is in discussions to purchase artificial‑intelligence startup Decart in a deal valued at roughly $7 billion. The WSJ noted that the talks could conclude soon, provided they do not collapse and no rival bidder emerges. PYMNTS reached out to Amazon for comment but had not received a reply at the time of publication.
Decart’s Technology and Founding Background
Founded in 2023, Decart has built software that makes it easier for developers to move workloads between different AI chips—a capability that has grown in importance as companies scramble for additional computing power to train and run large models. Beyond chip‑agnostic orchestration, the startup is also developing two world‑modeling systems named Lucy and Oasis, which can generate and simulate real‑world 3‑D environments. The WSJ highlighted that these tools could enable more immersive training scenarios for robotics, autonomous systems, and virtual‑world applications.
Amazon’s Trainium3 Chip and Competitive Landscape
The WSJ added that Amazon’s custom Trainium3 processor was created to compete directly with Nvidia, which continues to dominate the market for graphics processing units (GPUs) that power the current AI boom. Amazon Web Services (AWS) reportedly gave Decart early access to the Trainium3 chip last year after a meeting with the startup’s founders left AWS impressed by their vision. This early access suggests Amazon sees strategic value in integrating Decart’s chip‑portability software with its own silicon roadmap.
Comparative Deal Size and Amazon’s Recent Acquisitions
If the Decart transaction closes at the reported $7 billion, it would surpass Amazon’s $6.5 billion acquisition of MGM Studios in 2022 and rank among the company’s largest ever purchases. Earlier this year Amazon also announced an approximately $11 billion deal to acquire Starlink rival Globalstar, underscoring its willingness to make sizable bets across media, communications, and now AI infrastructure.
Prior Talks and Competing Interests: Anthropic and Calcalist Reports
Before Amazon’s entrance, Decart had been in advanced discussions with other suitors. The WSJ cited Israeli news outlet Calcalist, which reported recently that Decart was in advanced talks to be acquired by a major technology company. Additionally, the WSJ noted that AI safety‑focused firm Anthropic had walked away from a roughly $6 billion deal to purchase Decart last month, suggesting that price expectations or strategic fit may have shifted the negotiation landscape.
Amazon’s AI Push for Third‑Party Sellers: Seller Assistant Evolution
Beyond hardware acquisitions, Amazon is investing heavily in AI tools for its marketplace sellers. PYMNTS spoke last month with Mary Beth Westmoreland, Amazon’s vice president of Worldwide Selling Partner Experience, about the company’s next AI push for merchants. Westmoreland explained that the initiative is less about giving sellers better answers and more about reducing how often they need to ask questions in the first place.
Insights from Mary Beth Westmoreland on Seller Assistant Capabilities
In an exclusive conversation with PYMNTS, Westmoreland described how Amazon is expanding its Seller Assistant from a conversational tool into “something more akin to an operating layer for third‑party merchants.” She said, “We’ve gone from a smaller set of capabilities to onboarding just hundreds of them.” The assistant now does more than answer queries; it can reason across inventory, pricing, advertising, demand and compliance, remember how a merchant operates, continuously monitor the business, and take action when granted permission.
Implications for Sellers and Amazon’s AI Strategy
For Amazon’s millions of third‑party sellers, this evolution could transform AI from a peripheral toolkit item into the core infrastructure through which they run their stores. By automating routine decisions—such as repricing based on competitor moves, adjusting ad spend in response to demand signals, or flagging compliance issues—Seller Assistant aims to reduce operational friction and free merchants to focus on growth and product development. This aligns with Amazon’s broader AI strategy: layering proprietary silicon (Trainium3), flexible software (Decart’s chip‑agnostic layer), and intelligent services (Seller Assistant) to create an end‑to‑end AI stack that reinforces its cloud and e‑commerce dominance.
Conclusion and Outlook
The potential $7 billion acquisition of Decart, coupled with Amazon’s ongoing enhancements to Seller Assistant, signals a two‑pronged approach to strengthening its AI ecosystem. On the hardware side, securing Decart’s chip‑portability expertise could accelerate adoption of AWS’s Trainium3 and reduce reliance on Nvidia GPUs. On the services side, evolving Seller Assistant into an operating layer seeks to embed AI deeply into the daily workflows of third‑party merchants, thereby increasing platform stickiness and driving higher transaction volumes. While the deal remains subject to negotiation outcomes and competing offers, the moves underscore Amazon’s conviction that controlling both the silicon foundation and the intelligent application layer will be critical to maintaining its leadership in the rapidly evolving AI landscape.
For continued coverage of Amazon’s AI initiatives, subscribe to the PYMNTS daily AI Newsletter.
https://www.pymnts.com/news/artificial-intelligence/2026/amazon-weighs-7-billion-purchase-of-ai-startup-decart/