When OpenAI and Anthropic Go to Work for Public Health, Every Business Owner Should Pay Attention

When OpenAI and Anthropic Go to Work for Public Health, Every Business Owner Should Pay Attention

Nearly 40% of local health departments in the United States are not yet using AI at all. That statistic, cited by the Coalition for Health AI (CHAI), is the reason a landmark new programme called PULSE now exists. But the lessons it carries reach well beyond government corridors and into every small and mid-size business trying to figure out how to adopt AI responsibly and at scale.

PULSE, which stands for the Public Health Use Case and Learning Scaling Engine, is a structured pilot programme that will deploy generative AI tools from OpenAI and Anthropic across 10 state, local, tribal, or territorial public health jurisdictions. OpenAI and Anthropic have each donated 10 enterprise licences with capacity for up to 2,000 public health practitioners combined. Accenture is overseeing participant onboarding and will help develop reusable implementation playbooks based on what works and what does not. Pilots are scheduled to begin in autumn 2026, with published playbooks expected in 2027. CHAI's leadership council will select the participating jurisdictions and organize practitioners into five focus areas: biosurveillance and drug-wave prediction, social determinants of health mapping, operations and community-feedback analysis, public communications and multilingual translation, and automated clinical data retrieval using a FHIR query engine.

What makes PULSE significant is not just who is involved, but how the rollout is being structured. The programme explicitly references NIST's AI Risk Management Framework as a benchmark for evaluating AI systems according to their intended use, operating environment, affected parties, and potential consequences. CHAI's stated position is that governance and trust must come before scale. Dr. David Lakey, a former Texas health commissioner involved in the effort, framed it plainly: "Every major technological transformation succeeds or fails based on trust, governance and execution." Anthropic's head of beneficial deployments, Elizabeth Kelly, added that PULSE was designed to allow practitioners to test tools in their own environments with privacy, governance, and responsible-use measures incorporated from the start rather than bolted on afterward.

For small and mid-size business owners, this programme is a case study in how serious AI adoption actually works. The most important observation is not the technology itself but the structure surrounding it. PULSE does not simply hand practitioners a ChatGPT Enterprise login and wish them luck. It creates communities of practice, assigns use cases with defined scope, commits to producing documented playbooks, and grounds the whole effort in an established risk framework. That sequencing matters. When your business adopts AI tools for customer communications, lead generation, content production, or operations, the same architecture applies: define the use case first, then select the tool, then build the guardrails, then document what you learn.

The PULSE use case around public communications and multilingual translation is particularly instructive for business owners who serve diverse or multilingual customer bases. Using AI to translate outreach materials, respond to inquiries in multiple languages, or personalize messaging by community is not just a public health challenge. It is a local marketing challenge. The same tools being tested in government health departments are available to small businesses right now, and the opportunity to reach customers in their preferred language with consistent, accurate messaging is real and immediate. Similarly, the operations and community-feedback analysis use case mirrors what any business can do with AI-assisted customer sentiment analysis, reviewing inbound messages, support tickets, and survey responses to spot patterns that would take a team weeks to find manually.

The governance gaps highlighted in the PULSE announcement are also worth noting for business owners. The article is candid that CHAI has not yet published separate evaluation, privacy, security, or human-review requirements for each use case. It does not specify whether model outputs in biosurveillance or clinical retrieval will be reviewed by staff before being acted on. This transparency is valuable precisely because it names the open questions every AI deployment faces. For your business, the parallel questions are: Who on your team reviews AI-generated content before it goes to customers? What happens when the AI produces something inaccurate or off-brand? Do you have a documented policy? These are not hypothetical concerns. They are the practical governance questions that separate businesses that use AI well from those that run into costly mistakes.

This week, pick one workflow in your business where AI could save time or improve output, such as responding to customer inquiries, drafting marketing emails, or summarizing customer feedback, and write a one-page brief that defines the use case, names the tool you will test, identifies who reviews outputs before they are used, and sets a 30-day checkpoint to evaluate results. That is exactly the pilot-and-playbook model PULSE is using at a national scale, and it is the right way to start at any scale.

When government agencies and major AI labs structure their adoption the same way a disciplined small business should, that is not a coincidence. It is a signal that governance-first, use-case-specific AI strategy is the standard, and the businesses that build that discipline now will be the ones best positioned to scale it.

Originally inspired by: US public health agencies to test OpenAI and Anthropic AI models (https://www.artificialintelligence-news.com/news/openai-anthropic-public-health-ai/) See how Leads to Conversion can help you implement AI the right way from day one. Get your free AI audit