CDC's Vision for Using Artificial Intelligence in Public Health

About

CDC is committed to using artificial intelligence/machine learning for innovation, operational efficiency, and fighting infectious disease. CDC's artificial intelligence innovation approach includes investment areas, partnerships, workforce readiness, and guidance.

Artificial Intelligence

Vision

CDC staff and public health agencies nationwide will harness the abundance of opportunity that AI offers by safely and securely applying the tools available to them.

As a leader in AI, CDC wants to empower all staff to harness AI responsibly, streamline operations, and forge dynamic partnerships across industry, academia, other federal agencies, and state, tribal, local, and territorial public health agencies. By embracing the transformative power of AI, we are working to create a healthier future and improve the lives of all Americans.

Key Terms

Artificial intelligence (AI) is a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments. AI systems use machine- and human-based inputs to perceive real and virtual environments; abstract such perceptions into models through analysis in an automated manner; and use model inference to formulate options for information or action. (15 U.S.C. 9401(3)).


Machine learning (ML) means an application of artificial intelligence that is characterized by providing systems the ability to automatically learn and improve on the basis of data or experience (15 U.S.C. 9401(3)).

AI Innovation at CDC

Demonstrated Impact

  • Became the first federal agency to deploy a generative AI (GenAI) chatbot to all staff, contributing to over $3.7 millionAestimated in labor costs saved to date and a 527% return on investment.
  • Served as a source of GenAI guidance for more than 30 federal agencies.
  • Generated 55 AI solutions to challenges, or use cases, that demonstrate the power of AI in preventing outbreaks and enhancing operational efficiency. (Refer to the annual HHS AI Use Case Inventory.)
  • Developed an AI Accelerator, an internal program dedicated to operationalizing and scaling AI/ML technologies to solve complex public health problems.

AI/ML Use Cases

CDC is using AI to provide solutions to address specific challenges, referred to as use cases. CDC maintains an inventory of AI use cases through the annual HHS AI Use Case Inventory, in alignment with M-25-21, ensuring effective AI use in public health.

Analyzing Grant Reports with AI

Challenge: Reviewing 4,500 quarterly reports from recipients of a national grant program was a time-intensive process, requiring manual extraction of key insights from thousands of pages of unstructured text and data.

Solution: The CDC program team deployed AI-powered tools — CDC Chatbot with Microsoft's Azure OpenAI and Azure AI Search — to mine and analyze reports automatically, accelerating interpretation and improving reporting accuracy.

Impact: Enabled faster, more comprehensive analysis while improving reporting quality, reducing manual effort by an estimated 5,500 labor hoursB, and saving $500,000 in labor costsB.

Detecting Cooling Towers During Legionnaires' Disease Outbreaks

Challenge: Identifying cooling towers — potential sources of Legionella bacteria — during a Legionnaires' disease outbreak is critical but has traditionally required lengthy manual satellite image assessments.

Solution: Leverage AI to analyze satellite images and automatically detect cooling towers in affected regions, enabling rapid and accurate identification of potential outbreak sources.

Impact: Over 280 hoursC saved annually in investigative time to strengthen public health response efforts, mitigate the spread of Legionnaires' disease, and save lives by making faster intervention possible.

Analyzing News Articles to Support Outbreak Response

Challenge: Manual gathering and tagging of news data limit situational awareness for public health event monitoring, which can create delays in tracking outbreaks effectively.

Solution: Use AI to automate intake, categorizing, and summaries of thousands of news articles, providing rapid and scalable support for case-based and event-based surveillance.

Impact: Enhanced situational awareness, with approximately 8,000 articles processed a day, speeding up outbreak detection and increasing CDC's ability to monitor potential health threats.

Examples of CDC Programs Currently Using AI

National Syndromic Surveillance Program

Use of AI: The National Syndromic Surveillance Program uses AI for real-time analysis of patients' symptom data from emergency departments to detect outbreaks and monitor health trends. Machine learning algorithms help identify patterns that may indicate public health threats or disease trends.

Results: Improved detection of outbreaks, including faster response times and enhanced situational awareness during public health emergencies.

FluSight

Use of AI: Some forecasting teams submitting to FluSight use AI and ML to predict influenza — or flu — activity in the United States. These approaches can combine data from several sources like historical flu data and social media trends.

Results: More accurate flu forecasts can help public health officials, healthcare providers, and organizations better plan for the future and inform messages about anticipated flu increases.

How AI Supports CDC’s Public Health Data Strategy

CDC's Public Health Data Strategy (PHDS), launched in 2023 and updated each year with new milestones, supports swift, secure, and comprehensive exchange of health data. The agency will define and expand shared AI capabilities within its data platform in 2025, leveraging insights from 2024 applications.

AI plays a crucial role in accelerating the PHDS and strengthens all the annual milestones by:

  • Supporting response readiness by optimizing the exchange of critical health data.
  • Enabling rapid analysis of vast datasets, including images, audio, free text, and genomic information.
  • Identifying relationships in health data that traditional methods might overlook.

Building an AI-Ready Workforce

AI Accelerator (AIX)

CDC's AIX program is operationalizing and scaling AI/ML technologies for enterprise use and promoting the use of AI/ML across the agency.

The program prioritizes use cases that are significant to public health and ensures that AIX efforts align with CDC mission and goals. AIX is committed to creating safe and trustworthy AI/ML solutions while fostering innovative collaborative frameworks.

AI Community of Practice (CoP)

CDC's AI CoP brings together AI experts, enthusiasts, and practitioners from across the agency to share best practices and lessons learned in AI. The sessions feature presentations from internal teams and external partners with opportunities for collaboration.

In fiscal year 2024, CDC's AI CoP led monthly sessions for its more than 2,200 members including "CDC Chatbot 101," "Prompt Engineering," and the "Data Science Upskilling Program."

Working with Partners to Understand Needs and Support Innovation

CDC is working with public and private partners to drive adoption of AI and support innovation in the field.

Partnering with Public Health and Academia

To understand the needs of our nation's state, tribal, local, and territorial (STLT) public health agencies, CDC partnered with the CDC Foundation to assess awareness of, adoption of, and concerns about the use of AI/ML tools in health agencies. We learned that STLTs are looking for CDC guidance in two primary areas:

  • Pinpointing areas where AI can enhance public health operations.
  • Establishing strategies to be sure AI is deployed responsibly and securely.

Through collaboration with academic partners and state public health partners, CDC supports innovation in sharing public health data. We are:

  • Optimizing analytical capabilities – Breaking down data silos for more effective insights and decision-making.
  • Scaling AI for emergency response – Connecting fragmented data tools to improve scalability in crisis situations.
  • Improving public health monitoring – Advancing methods for tracking exposures and assessing the health of vulnerable communities.
  • Supporting workforce preparedness – Ensuring CDC personnel are equipped to address public health challenges using AI solutions.

Partnering with Private Industry

Harnessing advancements in AI technology holds immense promise for accelerating data-driven insights in public health. CDC is committed to regularly reviewing and integrating new technologies as they emerge to ensure timely, evidence-based insights for public health decision-making while maintaining robust human oversight, security, and research excellence.

AI/ML Authorities and Guidance

The White House and Office of Management and Budget have outlined authorities, policies, and guidance for AI use that are guiding CDC's AI efforts. CDC will continue to monitor and align efforts to reflect any new or updated direction from HHS, the White House, or other relevant authorities.

Federal Authorities, Policy, and Guidance

The White House

Office of Management and Budget Memoranda

Other Resources

Federal Resources

Resources for Public Health Agencies

Print
Content Source:
Office of Public Health Data, Surveillance, and Technology (OPHDST)
  1. Estimated calculations for return on investment, time saved, and cost savings are based on an internal, unpublished CDC analysis of chatbot usage to include: tokens, task types, industry benchmarks on time savings, estimated labor rates, implementation costs, infrastructure costs, platform costs, and training and adoption costs.
  2. Estimated costs and time-savings are based on an internal, unpublished CDC analysis.
  3. Estimated time-savings are based on an internal, unpublished CDC analysis.