New AI/ML solutions in AWS GovCloud (US) underpin responsible innovation

Technology leaders in the public sector want to explore artificial intelligence and machine learning (AI/ML) solutions to enhance mission impact and achieve business goals. At the same time, they are obligated to safeguard sensitive government data through alignment with strict regulatory compliance programs, standards, and Executive Orders such as 14110 requiring Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.

How can technology leaders rapidly deliver responsible AI-based innovation while aligning with the broad spectrum of strict regulatory requirements? The answer is simple: Amazon Web Services (AWS) GovCloud (US) provides the technology that underpins a solid foundation for securely and compliantly building and deploying AI capabilities.

AWS is a long-term AI/ML innovator in highly regulated hyper-scale cloud  

AWS realized more than a decade ago that technology leaders with US data residency, ITAR, FedRAMP High, and DoD SRG IL-4/5 compliance requirements needed a technology platform with AI/ML capabilities to promote innovation responsibly. Innovation like classifying images, conducting enterprise searches, processing and analyzing documents, transcribing audio, translating languages, and enabling computer-generated speech and conversational chat. AWS answered this need many years ago by deploying a suite of AI/ML capabilities, including Amazon SageMaker, Amazon Rekognition, Amazon Kendra, Amazon Textract, Amazon Transcribe, Amazon Translate, Amazon Polly, and Amazon Lex. This is why AWS added these AI capabilities to AWS GovCloud (US).

For years, AWS has believed that technology leaders need the power of choice when selecting AI/ML solutions. This is why AWS has focused on democratizing AI/ML technologies and promoting industry partnerships so that technology leaders using AWS GovCloud (US) can achieve a broad spectrum of mission and business benefits.

Generative AI innovation is the dawn of a new technology era 

The rapid pace of AI advancement in the early 2020s, coupled with 2023’s Executive Order 14110, elevated the visibility of the benefits of AI/ML solutions. Many technology leaders within highly regulated environments began 2024 by developing a strategy to responsibly take advantage of the vast opportunities to innovate with generative AI technologies. This is because they realized this new technology era provides innovative possibilities to elevate business and mission outcomes in ways that previously seemed impossible or uneconomical to achieve.

Shortly after Executive Order 14110, AWS introduced managed generative AI capabilities to AWS GovCloud (US) with the availability of Amazon Bedrock. Amazon Bedrock in AWS GovCloud (US) is now available with Guardrails for Amazon Bedrock (to help enable responsible AI practices) and the Anthropic Claude Sonnet 3.5 and Meta Llama 3 models. Amazon SageMaker Jumpstart  is now also available in AWS GovCloud (US), enabling the use of publicly available, open-weight models, sourced from Hugging Face and Meta. Bookmark our What’s new with AWS GovCloud (US) page for more updates.

The value of these AI solutions to technology leaders operating highly regulated environments is extreme: AWS GovCloud (US) customers can now responsibly advance their generative AI experimentation and development journey in a secure, resilient, sovereign, and isolated hyperscale cloud environment.

Showcase of mission outcomes using AI/ML 

AWS AI/ML capabilities, infrastructure, and compliance programs position technology leaders to responsibly deploy AI/ML using AWS GovCloud (US). Here are some examples:

Transform and enhance citizen experience — AI/ML-enabled Amazon Lex chatbots in webpages and contact centers powered by Amazon Connect can increase customer satisfaction and call velocity. Conversation analytics during customer interactions may be used to reduce time to resolution, improve customer experience, and create tailored customer experiences.

Elevate patient care quality and outcomes — AI/ML-enabled Amazon Textract and Amazon Comprehend underpin an AI/ML-based intelligent document processing (IDP) solution that will not only extract text and structured data from documents but will also employ AI/ML capabilities to provide business insights and determine relationships between patient datasets. The velocity of IDP helps healthcare providers efficiently manage patient care and automatically reduce the potential for information errors so providers can focus on positive patient health outcomes rather than on managing patient records and bills.

Improve employee experience and impact — AI/ML-enabled Amazon Kendra provides employers with solutions to help employees spend less time searching intranets and data repositories and opening countless documents to find answers to their questions. This results in increased productivity and decreased stress. AI/ML-enabled Amazon Transcribe and Amazon Comprehend enable employees to summarize meetings and dialogues through automated audio note collection, which saves employees time and improves information accuracy. This empowers leaders to make more informed business decisions and increases decision-making velocity.

These are just a few examples of how technology leaders can take advantage of the AI/ML capabilities available in AWS GovCloud (US) today—the possibilities are seemingly endless. How should technology leaders responsibly, compliantly, and securely start their AI/ML journey in AWS GovCloud (US)?

What should I do first? 

What can technology leaders do to dive into this fast-paced AI/ML era safely and responsibly? Here are three pieces of advice from the AWS GovCloud (US) team:

1. Start experimenting today, but don’t try to boil the ocean

Don’t make AI/ML adoption harder than it must be, or wait to get started. Start with a small yet well-defined AI/ML project with short-term measurable outcomes, broad executive support, and minimal risk. Demonstrate the success of this first project, and then kick off a slightly larger project with more impactful measures. Keep repeating this cycle in time-bound iterations and demonstrate increasing incremental mission impact using a “crawl, walk, run” approach. This is how Amazon started more than two decades ago when we launched our Amazon.com e-commerce recommendations engine underpinned with AI/ML technology. This AI/ML solution allowed us to elevate customer experience by providing a tailored shopping experience. Leaders can take similar approaches to address their unique business challenges, solve their most challenging problems, and responsibly innovate at scale—ultimately for the benefit of society. Technology leaders will benefit more from this approach than endless planning and waiting for the “perfect time” to start their journey.

2. Implement a data strategy that underpins a foundation of AI/ML innovation

Creating a robust enterprise data strategy is critical to implementing successful AI/ML solutions. Generative AI and foundation models (FMs) have elevated the need for clean training data. Quality and clean data are necessary to fine-tune generative AI models to unlock and maximize business and mission value. Technology leaders sometimes face situations where their data strategy does not provide quality training data, which causes the results from AI/ML solutions to produce inferior outcomes due to the “garbage in, garbage out” phenomena.

A well-defined data strategy will pay dividends to organizations embarking on their AI/ML journey. Whether you are building your model or customizing one, all leaders need a data strategy that ensures relevant, high-quality data is available. Some data may be more than 20 years old residing on a mainframe, while others might be massive unstructured datasets living in a legacy storage system. Data must be up-to-date, complete, accurate, discoverable, and available when needed. AWS can help accomplish these goals with Amazon Redshift for data warehouses, Amazon Simple Storage Service (Amazon S3) for data lakes, and Amazon EMR for big data. In turn, these services can increase the value of data as a component of an enterprise AI/ML solution.

3. Partner with AWS to promote positive, mission-driven outcomes

AWS offers our customers access to AI/ML-focused AWS GovCloud (US) specialist solutions architects and AI/ML-focused professional services. We also have a broad partner community of consultants and integrators. Visit the AWS AI/ML solutions library, download our public sector solutions guide to learn more, or visit AWS in the Public Sector to request assistance.

The bottom line on AI/ML and regulated cloud solutions 

The intersection of GPU-based cloud computing, AI/ML services, networking, and data storage with security and compliance enables highly regulated customers to confidently and responsibly deploy AI/ML solutions on AWS GovCloud (US). AWS encourages technology leaders to think big, experiment, responsibly innovate and take advantage of AWS GovCloud (US) as a strategic technology enabler to achieve their AI/ML-based mission outcomes.