AWS is already powering the Genesis Mission, a national initiative to accelerate American scientific discovery using AI and high-performance computing. Through its collaboration with Idaho National Laboratory and major investments in secure government cloud and AI infrastructure, AWS is helping researchers compress years of discovery into months—enabling breakthroughs in nuclear energy, national security, and advanced science.

AWS now offers regional availability for NAT Gateway, enabling a single managed gateway to automatically span multiple availability zones in a VPC. This update removes the need for zonal NAT gateways, public subnets, and complex route management—simplifying high-availability networking while maintaining built-in scaling and resiliency across AZs.

AWS has introduced a new Responsible AI Lens and updated its Machine Learning and Generative AI Lenses within the Well-Architected Framework. The expanded guidance helps architects and AI teams design, deploy, and govern AI systems with a focus on trust, fairness, security, and operational excellence—enabling organizations to scale AI responsibly on AWS.

AWS created its Generative AI Innovation Center to help customers turn AI pilots into production systems that deliver real business value. After working with more than 1,000 organizations, AWS has learned that success depends on leadership alignment, data readiness, ROI discipline, and employee adoption—along with growing demand for customized models and agentic AI as enterprises mature.

Amazon is transitioning from cloud pioneer to AI-first enterprise, embedding agentic systems across AWS, logistics, retail, and customer experiences. From Amazon Bedrock to AgentCore and consumer-facing agents like Alexa+ and Rufus, the company is showing how AI can drive real productivity gains while maintaining governance, security, and workforce transformation at scale.

AWS CEO Matt Garman explains why AI should empower developers rather than replace them, outlining AWS’s enterprise-first approach to agentic AI, custom model training with Nova Forge, and delivering measurable business impact. The conversation highlights AWS’s focus on practical innovation, workforce development, and building AI systems that integrate deeply into real-world enterprise workflows.

Chainalysis is now available on AWS Marketplace, allowing organizations to find, buy, and deploy blockchain intelligence solutions directly in their AWS environment. The launch includes tools for crypto compliance, investigations, and customizable data-driven threat intelligence, and is reinforced by Chainalysis joining the AWS ISV Accelerate Program—helping customers manage risk, reduce fraud, and support safer digital asset adoption.

Physical AI demands both millisecond responsiveness at the edge and deep reasoning in the cloud. AWS bridges this gap with Strands Agents, Amazon Bedrock AgentCore, and edge integrations that enable intelligent systems to perceive, plan, and act in real environments. From robotics to autonomous systems, AWS provides the foundation for building scalable physical AI from edge to cloud.

AWS has previewed Interconnect-multicloud, making it easier for enterprises to connect AWS environments with other major cloud platforms through high-speed private links. For teams pursuing best-of-breed cloud strategies, this simplifies cross-cloud networking and supports flexibility, resilience, and faster adoption of specialized services—while keeping AWS at the center of enterprise architecture.

Pinterest uses AWS to power one of the world’s largest AI-driven discovery platforms, processing over 500 PB of data to deliver 10 million personalized recommendations per second. Built on Amazon EKS, EC2, SageMaker, and Bedrock, Pinterest pairs scalable generative AI with a focus on safety, positivity, and responsible content moderation.