Amazon is aggressively investing in artificial intelligence (AI) to accelerate the expansion of its Amazon Web Services (AWS) business, according to President and CEO Andy Jassy. He highlighted the significant growth potential during a recent financial analyst call, despite ongoing supply chain constraints.
AWS Revenue Surges with AI Integration
Jassy reported that AWS revenue has grown by 17% year-over-year, reaching an annualized run rate of $117 billion. This growth is fueled by the increasing adoption of AI-based services among businesses. AWS continues to play a crucial role in helping organizations of all sizes transition to the cloud, modernize their infrastructure, reduce operational costs, and accelerate innovation.
The potential for further growth in the AWS sector is substantial. Jassy pointed out that over 85% of global IT spending is still directed towards on-premises infrastructure, indicating a significant opportunity for cloud adoption. He anticipates that this balance will shift dramatically over the next 10 to 20 years. Before the current advancements in AI, AWS was projected to become a multi-100-billion-dollar revenue business. However, with the integration of AI, Jassy believes that AWS has the potential to become even larger.
Aggressive Investment in AI Technologies
Amazon is making substantial investments in AI, driven by the belief that AI will transform customer experiences across various industries. Jassy emphasized that if a company’s mission is to improve customers’ lives daily, then investing aggressively in AI is essential. Amazon’s commitment to AI is evident in the development of over 1,000 AI applications across the company. This includes the next-generation Alexa, known as Alexa+, and the integration of AI in fulfillment networks, robotics, shopping experiences, Prime Video, and advertising.
AWS is also developing building blocks for both internal and external developers to create their own AI solutions. Jassy stressed that Amazon is not just experimenting with AI but is strategically providing developers with a broad range of capabilities at every level of the AI stack, ensuring cost-effectiveness and enabling widespread AI adoption across various businesses. The strategy involves not only research and development but also the practical application of AI technologies in various facets of its operations, from enhancing customer experiences to optimizing internal processes. This holistic approach underlines Amazon’s commitment to leveraging AI as a fundamental driver of growth and innovation.
Moreover, Amazon’s focus extends beyond merely creating AI tools; it emphasizes empowering developers to integrate AI into their own applications. By providing a comprehensive AI stack, AWS enables businesses to access sophisticated AI capabilities without the need for extensive in-house expertise. This democratization of AI is a crucial aspect of Amazon’s strategy, making AI accessible to a wider range of organizations and fostering a more innovative ecosystem.
The integration of AI also plays a key role in improving efficiency and reducing costs. By automating various tasks and processes, AI helps streamline operations and optimize resource allocation. This is particularly important in areas such as supply chain management, logistics, and customer service, where AI can significantly enhance performance and drive cost savings.
Trainium2 AI Chip: A Game Changer
At the foundational layer for model building, Amazon has introduced its new custom Trainium2 AI chip. This chip is experiencing significant demand due to its superior price performance. Jassy noted that while AWS supports multiple chip providers, customers operating AI at scale recognize that costs can escalate rapidly. The Trainium2 chip offers a 30% to 40% improvement in price performance compared to other GPU-based instances, making it an attractive option.
Reducing the cost of inference is crucial for the widespread success of AI. Amazon views this as part of its mission and responsibility. The Trainium2 chip is designed to help achieve this goal by lowering the overall cost of AI inference, making it more accessible for various applications. This is particularly significant for companies deploying large-scale AI models, where inference costs can be a major barrier to adoption. By optimizing the performance of its AI chips, Amazon is not only enhancing its own capabilities but also contributing to the broader accessibility and affordability of AI technologies.
The Trainium2 chip also highlights Amazon’s commitment to innovation in hardware design. By developing its own custom chips, Amazon can tailor the hardware specifically to the needs of its AI workloads, resulting in superior performance and efficiency. This vertical integration of hardware and software allows Amazon to optimize the entire AI stack, delivering greater value to its customers. Furthermore, the development of custom AI chips demonstrates Amazon’s long-term vision for AI, positioning itself as a leader in both AI software and hardware.
Amazon Bedrock: Simplifying GenAI Application Development
In the middle layer, AWS offers Amazon Bedrock, a managed service designed to simplify the development of GenAI applications. Bedrock provides customers with a choice of high-performing foundation models and a comprehensive set of features that facilitate the creation of high-quality generative AI applications. This approach allows businesses to leverage the power of generative AI without the complexity of managing the underlying infrastructure. By abstracting away the technical details, Bedrock enables developers to focus on building innovative applications that meet their specific needs.
Amazon continues to enhance Bedrock by adding advanced models such as Anthropic’s Claude 3.7 Sonnet hybrid reasoning model and Meta’s Llama 4 family of models. AWS was also the first cloud service provider to offer DeepSeek R1 and Mistral AI’s Pixtral Large as fully managed models, further expanding the options available to developers. The continuous addition of new models ensures that Bedrock remains at the forefront of generative AI technology, providing customers with access to the latest and most advanced capabilities. This commitment to innovation is crucial for maintaining a competitive edge in the rapidly evolving AI landscape.
The availability of a diverse range of foundation models is a key differentiator for Bedrock. By offering models from various providers, AWS allows customers to choose the models that best suit their specific use cases. This flexibility is particularly important in the field of generative AI, where different models excel at different tasks. Whether it’s generating text, images, or code, Bedrock provides a wide array of options to cater to the diverse needs of its customers.
Introducing Amazon Nova: Frontier Intelligence
Amazon recently introduced the latest version of its Amazon Nova foundation model with Bedrock. These models deliver frontier intelligence and industry-leading price performance. Thousands of customers are already utilizing them for various AI applications. Amazon also released Amazon Nova Sonic, a new speech-to-speech foundation model that enables developers to create highly accurate, expressive, and human-like voice-based AI applications. Nova Sonic offers lower word error rates and higher win rates compared to other comparable models for speech interactions.
The introduction of Amazon Nova highlights Amazon’s dedication to pushing the boundaries of AI technology. By developing its own foundation models, Amazon can ensure that its AI services are tailored to the specific needs of its customers. This also allows Amazon to maintain greater control over the AI stack, optimizing performance and security. The focus on frontier intelligence underscores Amazon’s commitment to delivering cutting-edge AI capabilities that can transform various industries.
The release of Amazon Nova Sonic further demonstrates Amazon’s focus on innovation in specific areas of AI. By developing a dedicated speech-to-speech model, Amazon is addressing the growing demand for voice-based AI applications. The emphasis on accuracy and expressiveness reflects Amazon’s commitment to delivering high-quality user experiences. With its lower word error rates and higher win rates, Nova Sonic sets a new standard for speech interaction, enabling developers to create more natural and engaging voice-based AI applications.
Investing in Agentic AI
Amazon is heavily investing in agentic AI, moving beyond traditional question-answer applications. The goal is to develop agents capable of performing complex, multi-step tasks, such as organizing trips, adjusting lighting and temperature, setting music ambience for dinner guests, and handling complex IT tasks to improve business productivity. The first action-oriented agents are being developed with Alexa+, although Jassy acknowledged that the technology for building these agents is still in its early stages and requires ongoing human supervision.
The shift towards agentic AI represents a significant evolution in the field of artificial intelligence. By enabling AI systems to perform complex tasks autonomously, agentic AI has the potential to revolutionize various industries. From automating routine tasks to assisting with complex decision-making, agentic AI can significantly improve efficiency and productivity. Amazon’s investment in this area reflects its belief that agentic AI will play a key role in the future of AI.
The development of action-oriented agents with Alexa+ is a strategic move by Amazon. By leveraging its existing voice assistant platform, Amazon can rapidly deploy agentic AI capabilities to a large user base. This also allows Amazon to gather valuable data and feedback, which can be used to further improve the performance of its agents. While the technology is still in its early stages, the potential benefits of agentic AI are immense, and Amazon is well-positioned to capitalize on this emerging trend.
Amazon recently released a research preview of Amazon Nova Act, an AI model trained to perform actions within a web browser. This model enables developers to break down complex workflows into reliable atomic commands, such as “search” or “checkout,” and answer questions about the screen. It also allows for the addition of detailed instructions to these commands, such as “don’t accept the insurance upsell.” Nova Act aims to improve the accuracy of multi-step agentic actions from the current state of 30% to 60% to over 90%, providing the necessary building blocks for creating action-oriented agents.
Amazon Q: Accelerating Software Development
At the top of the AI stack is Amazon Q, a GenAI-powered assistant designed to accelerate software development. It leverages a customer’s own data and provides a fast, new agent coding experience within the command line interface to autonomously execute complex workflows. Amazon has also made GitLab Duo with Amazon Q generally available, allowing AI agents to assist with multi-step tasks.
Amazon Q represents a significant step forward in the application of AI to software development. By leveraging generative AI, Amazon Q can assist developers with various tasks, such as code generation, bug fixing, and documentation. This can significantly improve developer productivity and accelerate the software development process. The integration with GitLab Duo further enhances the capabilities of Amazon Q, allowing AI agents to assist with multi-step tasks within the GitLab environment.
The use of customer’s own data is a key differentiator for Amazon Q. By leveraging the specific data and context of each customer, Amazon Q can provide more relevant and accurate assistance. This also helps to ensure that the generated code and documentation are tailored to the specific needs of the customer. The combination of generative AI and customer-specific data makes Amazon Q a powerful tool for accelerating software development.
Amazon’s AI business has achieved a multi-billion-dollar annual revenue run rate and continues to grow at triple-digit year-over-year percentages. Despite the optimism surrounding AI, Jassy reminded stakeholders that a significant amount of on-premises infrastructure still needs to be moved to the cloud. Infrastructure modernization is crucial for any company’s technology, innovation capabilities, developer productivity, speed, and cost structure. Companies must move their infrastructure and data to the cloud to fully realize the potential of AI.
The success of Amazon’s AI business underscores the growing demand for AI-powered services. As AI technology continues to advance, more and more businesses are recognizing the potential benefits of integrating AI into their operations. Amazon’s comprehensive AI stack, which includes custom chips, foundation models, and AI-powered assistants, provides a complete solution for businesses looking to leverage the power of AI. The reminder about the importance of cloud migration highlights the crucial role that cloud infrastructure plays in enabling AI.
Addressing Supply Constraints
During the question-and-answer session, Jassy addressed concerns about supply constraints related to chips for AI. He stated that AWS’s growing multi-billion-dollar business means that it is constantly adding capacity and utilizing new chips as soon as they become available.
Jassy believes that AWS could be helping more customers and generating more revenue if it had access to more capacity. He noted that more Trainium2 instances and next-generation Nvidia instances are expected to arrive in the coming months. Other components of the supply chain, such as motherboards, are also experiencing delays due to high demand. However, Jassy expects that these supply chain and capacity issues will continue to improve throughout the year.
The supply constraints related to AI chips highlight the growing demand for AI hardware. As AI technology becomes more prevalent, the demand for high-performance chips that can power AI workloads is increasing rapidly. Amazon’s efforts to address these supply constraints demonstrate its commitment to ensuring that it has the necessary resources to meet the growing demand for its AI services. The expectation that supply chain and capacity issues will improve throughout the year provides some optimism for the future.
Impact of Tariffs
Jassy also addressed the impact of the Trump administration’s tariffs on Amazon’s retail sales. He mentioned that there has been some increased buying in certain categories, potentially indicating stocking up in anticipation of tariff impacts. However, the average selling price of retail items has not significantly increased yet. This is partly due to forward buying by Amazon and advanced inbounding by third-party sellers. Many sellers have not yet adjusted their pricing.
The situation could change depending on where tariffs ultimately settle. Jassy stated that Amazon is not uniquely susceptible to tariffs related to China. Retailers who do not buy directly from China typically purchase from companies that do, marking up the items, rebranding them, and selling them to U.S. consumers. These retailers are buying the product at a higher price than Chinese sellers selling directly to U.S. consumers in Amazon’s marketplace, resulting in higher overall tariffs for these retailers and for China direct sellers.
Thediscussion of tariffs highlights the complex interplay between global trade and retail sales. The potential impact of tariffs on consumer prices and business operations is a significant concern for retailers like Amazon. The strategies employed by Amazon and its third-party sellers to mitigate the impact of tariffs, such as forward buying and advanced inbounding, demonstrate the adaptability of businesses in response to changing trade policies.
Financial Results
For the first fiscal quarter of 2025, which ended on March 31, Amazon reported total revenue of $155.7 billion, a 9% increase compared to the $145.3 billion reported for the first fiscal quarter of 2024.
This included:
- AWS sales of $29.3 billion, up 17%
- North America sales of $92.9 billion, up 8%
- International sales of $33.5 billion, up 8%
Total revenue exceeded analyst expectations by $580 million, according to Seeking Alpha.
Amazon also reported total GAAP net income of $17.1 billion, or $1.59 per share, up from last year’s $10.4 billion, or 98 cents per share. The company also reported operating income of $18.4 billion, up from last year’s $15.3 billion.
The strong financial results reported by Amazon for the first fiscal quarter of 2025 reflect the company’s continued growth and profitability. The impressive growth of AWS sales, coupled with solid performance in North America and international markets, demonstrates the strength and diversity of Amazon’s business. The fact that total revenue exceeded analyst expectations further underscores the positive trajectory of the company.
The significant increase in net income and operating income is a testament to Amazon’s ability to effectively manage its costs and improve its operational efficiency. These strong financial results provide a solid foundation for Amazon’s continued investment in AI and other strategic initiatives.
Future Outlook
Looking ahead, Amazon expects second fiscal quarter 2025 sales to be between $159.0 billion and $164 billion, representing growth of 7% to 11% over the second fiscal quarter of 2024.
The company also anticipates operating income to be between $13 billion and $17.5 billion, compared to $14.7 billion in the previous year.
The future outlook provided by Amazon indicates continued growth and profitability in the coming quarter. The projected sales growth of 7% to 11% reflects the company’s confidence in its ability to maintain its momentum. The anticipated operating income of $13 billion to $17.5 billion further reinforces this positive outlook. These projections provide a sense of optimism for the future of Amazon and its continued success in the AI and cloud computing markets. The strategic investments in AI and cloud infrastructure are expected to drive further growth and innovation in the years to come.