By Raphael Satter and Humeyra Pamuk
WASHINGTON (Reuters) -The U.S. State Department will use an artificial intelligence chatbot to help it select the people who will perform annual reviews of promotions and moves, according to a cable issued Monday and reviewed by Reuters.
The cable said that StateChat, an in-house chatbot which works using technology from Palantir and Microsoft, will be employed to pick foreign service officers for participation on the Foreign Service Selection Boards, the annual evaluation panels which decide whether and how to promote and shuffle around State Department employees.
In a statement, a department spokesperson said the evaluations themselves “will not be done by AI.”
The boards, whose role is governed by the 1980 Foreign Service Act, play a critical role in the State Department’s personnel promotion decisions, managing the annual process by which diplomats and others jump from one professional grade to the next. By statute, the boards are meant to include “a substantial number of women and members of minority groups.”
The State Department has been using StateChat since last year to transcribe notes, draft emails, and analyze diplomatic cables. Last week the agency’s acting chief data and AI officer, Amy Ritualo, told a Palantir conference that StateChat had about 40,000 users across her agency. The program’s role in the human resources process, however, has not previously been disclosed.
Last month the State Department abruptly postponed the boards, and previously selected members received emails saying their services were no longer required.
Monday’s cable said that StateChat’s technology would instead be used to “perform unbiased selection” for the boards based on employees’ internally adjudicated skill codes and grades. That list would then be screened – for example for disciplinary and security issues – before being used to create the panels. There was no mention of female or minority representation.
President Donald Trump’s administration has repeatedly attacked what Republicans refer to as “DEI,” a catch-all term covering work protecting civil rights, fighting discrimination, and boosting diversity.
The American Foreign Service Association, which represents State Department employees, did not directly comment on the use of AI but said it was seeking clarification from agency leadership about how it intends to comply with its legal obligations around women and minority group representation.
Palantir and Microsoft didn’t immediately return messages.
Although the deployment of AI by officials precedes Trump’s reelection in 2024, his administration has aggressively expanded its use since his return to power. Last month Reuters reported that tech tycoon Elon Musk’s U.S. DOGE Service was expanding its use of the AI chatbot Grok across the U.S. federal government.
In April, Reuters reported that Trump administration officials had told some U.S. government employees that DOGE was using AI to monitor at least one federal agency’s communications for hostility to the president.
(Reporting by Raphael Satter and Humeyra Pamuk; editing by Edward Tobin)