15 things the State Department said about the future of PEPFAR and global health

At AIDS 2026, U.S. officials offered their clearest account yet of how the new bilateral health agreements will work.

Senior State Department officials SBO Jeff Graham, DAS Dr. Rebecca Bunnell, DAS Dr. Mamadi Yilla and Special Advisor Carol O’Connell came to AIDS 2026 to explain the implementation of the administration’s new global health strategy and bilateral memoranda of understanding.

I had previously identified senior State Department officials among my 10 people to watch at AIDS 2026. It was noteworthy—and important to recognize—that they showed up in Rio to explain their approach and engage directly with the HIV community. Showing up does not resolve the disagreements, but it creates an opportunity for scrutiny, accountability and dialogue.

Activists interrupted the session with chants of “Restore PEPFAR now,” challenging the administration’s entire global HIV strategy over the past year. Their intervention ([see their statement here]) forced the room to confront two urgent issues: the administration’s continued withholding of congressionally appropriated PEPFAR funding, which they described as unlawful, and mounting evidence of deteriorating clinical outcomes.

Once the presentation resumed, officials made several important announcements and policy statements. Here are the 15 that stood out to me.

1. “We are changing the model, not the mission.”

This was Jeff Graham’s central message. Officials said the United States remains committed to ending HIV but is replacing donor-managed programs with government-led systems.

The question, of course, is whether the model can change without disrupting the mission—or the people who depend on it.

2. State now has 34 bilateral agreements involving major financial commitments.

Officials said 34 agreements are in place. The United States has committed nearly $15 billion over five years, while partner governments have pledged approximately $10.2 billion.

State described the co-financing commitments as “carefully measured” against each country’s fiscal circumstances. A country-level breakdown will be essential to determine how much of the $10.2 billion represents new and additional health spending.

3. State says only 40 cents of every assistance dollar previously reached the front line.

Officials claimed that 60 cents of every dollar went to overhead, technical assistance and program management, leaving only 40 cents for frontline health workers and commodities.

That is a striking statistic—and one that deserves a published methodology. “Overhead” can conceal inefficiency, but it can also include laboratories, surveillance, training, quality assurance and other functions that keep services working.

4. Parallel PEPFAR infrastructure will be reintegrated into national systems.

State said separate supply chains, laboratories, data platforms and disease-specific workforces are inefficient and unsustainable.

The goal is to integrate them into government systems. That makes sense in principle. The risk is dismantling functioning infrastructure before national systems have the financing and capacity to absorb it.

Importantly, the State Department commissioned the National Academies to produce a rapid expert consultation on transitioning disease-specific programs into integrated, country-led health systems. I served as a reviewer. The report examines evidence and implementation considerations involving governance, financing, workforce, data, laboratories and surveillance.

It is notable that State invested in independent evidence generation during such a consequential transition. I hope State and partner governments now use that evidence to guide implementation.

5. The five-year MOU period is not necessarily a five-year transition deadline.

Officials acknowledged that they do not know how long every transition will take. Secretary Rubio has previously suggested that some countries may require seven, 10 or even 15 years.

That flexibility matters. Disease burden, debt, fiscal capacity, political instability and health-system readiness cannot be squeezed into one standardized timeline.

6. State is holding back funds for central management.

Not all funding will flow through the bilateral agreements. State will retain several central funding “buckets” for:

• Innovation

• Performance incentives

• Transition resilience and unexpected shocks

Under innovation, officials highlighted the partnership with Gilead to expand access to lenacapavir, Zipline’s last-mile delivery model and new malaria technologies (slide below). They distinguished between product innovation and process innovation—both of which could change how countries prevent disease and deliver services.

Officials also said countries that exceed agreed targets could receive additional funding. The design, size and governance of these funds remain important unanswered questions.

7. The new APS funding mechanism is intended to bring in different partners.

State described its Annual Program Statement as a way to broaden the partner base beyond large international implementers.

Officials clarified that “quite a few” countries may receive direct government-to-government financing. Where that is not yet feasible, State will continue funding implementing partners (including some international NGOs) but with greater emphasis on local faith-based organizations, domestic provate sector and other local partners.

Decisions about implementing partners will be made jointly, although each government will retain final authority over how its own taxpayer funds are spent.

Localization could improve sustainability. But local ownership should not become code for outsourcing public responsibilities without adequate oversight.

8. State wants health data systems to be locally operated.

Officials said PEPFAR data systems briefly went offline during the past year, exposing the risks of externally operated platforms.

The new approach will shift systems toward country ownership because State believes local systems will be cheaper, more sustainable and more accessible to governments. The real test will be whether they remain timely, interoperable and capable of protecting HIV-specific visibility.

9. Aggregate data sharing will continue; specimen sharing applies only in some countries.

State said countries will still share epidemiological and financial data needed to track program targets and co-financing. Officials stressed that this does not include identifiable medical records.

They said specimen-sharing provisions appear in only a few agreements and are intended to facilitate rapid responses to disease outbreaks. These assurances should be compared with the actual legal language of each MOU.

10. State says civil society is part of implementation.

Officials said civil society will help monitor continuity, validate government data and identify communities being left behind. Some plans may include social contracting to finance community-led services.

While it was good to hear that emphasized out loud, affected communities need a seat at the table before decisions are finalized—not merely an invitation to monitor the consequences (which are also important).

11. The proposed implementation plans contain country-specific strategies for impact.

Dr. Rebecca Bunnell emphasized that the implementation plans are not meant to be financing documents alone. They reportedly contain approaches covering integration, prevention, diagnostics, technology, service delivery and health-system strengthening.

That is encouraging. These plans should all be published (as has been done previously for 20 years) so communities, researchers and legislators can assess whether the proposed changes are epidemiologically sound.

12. Triple elimination is emerging as a major integration strategy.

State highlighted efforts to address HIV, syphilis and hepatitis B together during pregnancy. Officials said approximately 120,000 children still acquire HIV each year and that 20 African implementation plans include the use of triple diagnostic tests.

This is the kind of integration that could expand impact while protecting HIV outcomes—provided countries maintain the commodities, workforce and follow-up systems needed to deliver it.

13. State views treatment as the highest-value HIV prevention intervention—but says prevention must include everyone at risk.

Officials argued that diagnosing people with HIV, initiating treatment and achieving viral suppression remain the most effective prevention investments.

They said PrEP should focus on people at the highest risk and that programs should consider the number of people who must receive PrEP to avert one infection.

Asked how prevention services would be intentionally designed for key populations, Dr. Rebecca Bunnell said the epidemic cannot be ended without including all people at risk. This was good to hear out loud!

That creates a significant accountability gap. A commitment to include everyone is important, but without specific targets and data, it will be difficult to determine whether key populations are actually being reached—or quietly left behind.

14. AI is already appearing in country implementation plans.

I was glad to see technology move beyond the buzzword stage.

State highlighted El Salvador’s use of community telehealth and Uganda’s deployment of portable, AI-supported chest X-ray systems for tuberculosis detection. Uganda reportedly plans to expand from 14 systems to 150.

These examples show where AI may add practical value: extending access, supporting health workers and finding cases earlier—not replacing health systems with chatbots and crossed fingers.

15. The future of HIV service delivery, State said, is simplification.

Officials repeatedly emphasized modernizing and simplifying service delivery. They pointed to differentiated care, less burdensome monitoring for stable patients and more efficient diagnostic and delivery systems.

They also highlighted self-care and new service-delivery models for both prevention and treatment. These approaches could include HIV self-testing, telehealth, community- and pharmacy-based services, longer medication refills and care models that require fewer routine clinic visits. The broader goal is to bring services closer to people while reducing costs and pressure on health facilities.

There will be a lot of evidence presented at AIDS 2026 on self-care, demedicalizing PrEP, differentiated treatment and new community-based delivery models. These presentations will be important to watch because they could directly inform the country implementation plans now being finalized.

Simplification should not become a euphemism for offering fewer services. Done well, it can make prevention and treatment easier to access, less medicalized and more responsive to people’s lives. The challenge is moving the evidence from conference rooms into national policy and implementation.

My bottom line

I left the session with a clearer understanding of how State sees the new model working and with a few unanswered questions. What happens in Zimbabwe, Zambia and South Africa?

There are sensible ideas here. PEPFAR has accumulated inefficiencies and parallel systems over 23 years. Country ownership cannot remain something we talk about but never achieve.

But people do not experience an aid transition through MOUs, financing formulas or organizational charts. They experience it through whether their clinic stays open, their medicines remain available and someone they trust is still there to help.

We should all want this transition to succeed. The test is simple: Are people staying on treatment? Is prevention reaching those who need it? Are new infections falling? Are communities still shaping the response?

Ultimately, that is what saves lives.