AWS' advertising offerings continue to grow, while newer businesses like Prime Video, Pharmacy, and Grocery provide different customer experiences.

Jassy wrote about being on the verge of launching Amazon Leo,  high-speed internet access, and how the company is just beginning to offer Zoox, its autonomous-driving ride-hailing service.

For the third consecutive year, Jassy used his annual shareholder letter to outline Amazon’s vision for AI as a key driver for growth.

This time he explained how Amazon would reinvent itself to support change, giving priority to a faster delivery service and high-speed internet access to rural communities, and making it easier for brands to reach consumers. One way to do this is through Alexa+.

Jassy pointed to the outcome from Amazon reinventing Alexa, which now has around 600 points in which consumers can access the platform, from cars to Prime Video.

Alexa has become one of Amazon's most interesting performance advertising plays by serving consumers with high intention lower in the advertising and marketing funnel. Adding to the cart for now or later, and then making a purchase are all moments that can be measured and tracked.

"We had to completely rewire her brain, corresponding intelligence, breadth of knowledge, routing of services and APIs she accesses, and what routines and jobs she could do," Jassy wrote. This was all done by going back to the start to begin again, but not completely from scratch.

It has been worth it, he wrote, because users are speaking with Alexa twice as much and for longer duration across a wider breadth of topics. They are completing purchases on devices three times more, streaming music 25% more, and using smart-home functionality 50% more.

Amazon's investments in AI also drive innovation in other companies. Take the NFL, for example. With the 2026 NFL draft approaching, fans are driven into mock drafts, prospect rankings, and team assessments. The challenge is not finding information, but rather in knowing what information matters and how it all connects.

Skai, a commerce media platform, is another example. On Thursday, Skai announced a new capability, Model Context Protocol (MCP), which enables agents to operate across all channels in real time, including Amazon as AI shifts from experimental to execution.  

Since many platforms are not designed for this shift, data remains fragmented, inconsistent, and difficult for AI systems to use in real-time.

Skai addressed this by building a foundation that looks at data as a human and machine workflow. More than half of internet traffic will come from machines as agents traverse the underbelly looking for products, prices and more to share with potential consumers.

Skai MCP's AI agent, Celeste -- or one used by another company -- can securely access, analyze, and act on this data in real-time without requiring custom APIs or complex integrations.

It is amazing to consider that Celeste has been used by more than 1,300 users across 170 organizations to analyze performance, identify opportunities, and automate workflows across 300 publishers and retail media networks.

MCP extends this capability beyond one agent, opening the platform to an ecosystem of AI systems.

These are the type of examples Jassy wrote about in his letter to shareholders, although he did not use this one. 

"AI is not a standalone initiative, it’s a multiplier," Jassy wrote. "It will reshape every customer experience we offer and unlock entirely new ones. We will build many of these ourselves, and continue making AWS the best place for others to do the same."