US apologises after AI-generated map wrongly represents African countries at global conference: List of countries, everything that happened

• The U.S. State Department has apologised after an AI-generated map incorrectly labelled and misplaced several African countries during a presentation at the AIDS 2026 conference in Brazil.

• The erroneous map wrongly showed Nigeria as a landlocked country and placed Côte d'Ivoire in East Africa, prompting widespread criticism from diplomats, experts and politicians.

• The State Department accepted full responsibility for the error, describing it as an unfortunate mistake made while updating presentation slides and reaffirming its commitment to working with African partners.

Artificial intelligence is increasingly being used to create presentations, reports and images, but one AI-generated map has landed the U.S. State Department in an embarrassing controversy after it wrongly labelled and misplaced several African countries during an international HIV/AIDS conference, according to reports from BBC.

The mistake drew criticism from netizens, researchers and politicians, forcing the state department to clarify.

Here's what happened and why it matters.

The incident occurred during the 26th International AIDS Conference (AIDS 2026) in Rio de Janeiro, Brazil, where a senior U.S. official delivered a presentation on funding opportunities under the President's Emergency Plan for AIDS Relief (PEPFAR).

One of the presentation slides contained a map of Africa that incorrectly identified and positioned several countries.

Images of the slide drew criticism after HIV/AIDS expert Emily Bass shared a photograph from the conference. Reuters later reported that the image carried an AI watermark, indicating it had been created using artificial intelligence tools.

The map contained several major geographical mistakes, including:

• Nigeria was shown as a landlocked country instead of being located along the Gulf of Guinea in West Africa.

• Côte d'Ivoire (Ivory Coast) was wrongly placed in East Africa instead of West Africa.

• Mozambique appeared in the wrong part of the continent.

• Malawi and Uganda were roughly in the correct region but appeared distorted.

• Cameroon was labelled on the slide, but its name was not correctly attached to any country.

Many critics said the errors went beyond a simple typo. The presentation was delivered at a major global health conference attended by African governments, health experts and international organisations. Critics argued that such basic geographical mistakes undermined confidence and reflected poorly on a government presentation involving African partners.

Former U.S. State Department official Cameron Hudson described the mistake as "stupid and embarrassing," saying it reflected poor oversight.

U.S. Senator Jeanne Shaheen also criticised the incident, arguing that relying on AI without proper human review can lead to serious mistakes.

The U.S. State Department accepted responsibility for the error. In a statement quoted by BBC, the department said the incident was "an unfortunate error" caused after a team member hurriedly modified the presentation shortly before the conference.

The department also said it took "full responsibility for the confusion and misrepresentation it caused for attendees, including our African partners."

According to an investigation by Reuters, the presentation slide contained an AI watermark, suggesting AI tools were used in creating or editing the image. However, neither Reuters nor the U.S. State Department said AI alone caused the mistake.

Instead, the department said the slide had been hurriedly altered before the event. Reuters reported that OpenAI is investigating the incident.

The State Department said the mapping error does not change America's commitment to global HIV/AIDS efforts but remains "substantive and constructive" reaffirming Washington's commitment to working with African countries through PEPFAR and other health partnerships.

The incident has become a reminder that while AI can speed up content creation, it can also produce significant errors if its output is not carefully checked.