National security officials at one of the world’s top artificial intelligence labs discussed “offensive” military uses of their technology for Ukraine in a prank call.
asserted
officials → discuss → call
Three senior OpenAI officials, including national security policy head Sasha Baker, strategic initiatives lead Caroline Zier and European national security lead Ben Crampton, participated in a 17-minute video call with two famous Russian pranksters impersonating Ukraine’s foreign intelligence chief Rustem Umerov, according to Vovan and Lexus’ video “Prank with AI Management” posted Wednesday on Rumble.
uncertain
officials → include → Rumble
The impersonators, Vladimir Kuznetsov (Vovan) and Alexei Stolyarov (Lexus), convinced the officials to discuss the company’s interest in contributing to the Ukraine conflict.
asserted
impersonators → convince → conflict
Baker told the pranksters that the company had a “very new” arrangement with the Department of War (DOW), according to the video call.
uncertain
company → tell → call
(RELATED: Ex-Army Boss Who Put ChatGPT Execs In Uniform Lands At OpenAI)
asserted
Who → Put → OpenAI
Vovan and Lexus asked the officials if they would give Ukraine access to AI models to support its military efforts, according to the call.
uncertain
they → ask → call
“There’s no question, I think, from our perspective that we want to have as deep a relationship as we possibly can with Ukraine,” Crampton told the pranksters.
uncertain
Crampton → ’ → pranksters
“But we have to go through a process … to get there.”
asserted
we → have → process
“There’s a general reluctance on the part of many people who are senior in the company to get too involved in offensive uses of our technology in a military context,” Crampton explained to the pranksters.
asserted
Crampton → ’ → pranksters
“Part of our job is to explain that offense and defense are not necessarily distinguishable from each other under certain circumstances and that we have to think about how we can help countries in defense.”
asserted
we → explain → defense
OpenAI announced a $200 million contract with the DOW in June 2025.
asserted
OpenAI → announce → June
The contract “will … identify and prototype how frontier AI can transform its administrative operations,” the company said in the announcement.
asserted
company → identify → announcement
The success of the project, called “Frontier AI Model Prototype,” will be judged by whether “warfighters, commanders” and others confirm “the capability would improve military advantage, military utility, or enhance military decision making,” according to a Feb. 27 amendment to the contract, obtained by the Daily Caller News Foundation.
“The overwhelming majority of the government and our researchers strongly support Ukraine,” Baker told the prankster duo during the call.
uncertain
Baker → call → call
‘Prank Journalists’
Vovan and Lexus said the call took weeks to coordinate.
asserted
call → say → weeks
To get in contact with OpenAI, the duo impersonated Umerov and reached out to a Ukrainian official in San Francisco who was in contact with Baker.
asserted
who → get → Baker
Baker served as deputy under secretary of defense for policy from Feb. 2022 through April 2024, according to the DOW’s official biography and a statement from then-Defense Secretary Lloyd Austin.
uncertain
Baker → serve → Austin
She has kept up her connections with Ukrainian officials following her departure from the DOW, Vovan and Lexus told the DCNF.
asserted
Vovan → keep → DCNF
The Ukrainian individual then set up their meeting with the AI company to “help lobby for Ukraine to gain access not only to standard civilian technologies, but also to frontier AI models and AI technologies intended for military use,” Vovan and Lexus told the DCNF.
asserted
Vovan → set → DCNF
To prepare for the call, the Ukrainian official sent the duo a briefing document, written in Ukrainian, that outlined concerns that China could restrict the country’s access to its open-weight models, the pair said.
uncertain
pair → prepare → models
Vovan and Lexus also asked OpenAI’s ChatGPT for advice on what questions they should pose to the company’s officials prior to the video call, and the chatbot said it would be interesting to see what the officials would tell them privately, the duo told the DCNF.
asserted
duo → ask → DCNF
“We call ourselves ‘prank journalists,’” the pair told the DCNF.
asserted
pair → call → ourselves
“We see our work as a form of investigative journalism conducted through pranks.”
asserted
We → see → pranks
“Our goal is not only to entertain people, but also to uncover information that would otherwise remain behind closed doors,” Vovan and Lexus told the DCNF.
asserted
Vovan → entertain → DCNF
“In this particular case, we wanted to understand just how far cooperation between AI companies and governments can go, and how quickly civilian partnerships can evolve into military ones.”
asserted
partnerships → want → ones
The pair believes the public has a right to know how AI is being used in military efforts, they told the DCNF.
asserted
they → believe → DCNF
Vovan and Lexus noted that OpenAI employees have “differing views” about how AI should be used for offensive capabilities and that “some people inside the company recognize the dangers of crossing that line.”
asserted
people → note → line
The AI lab said it investigated the call and took steps to prevent future incidents.
asserted
it → say → incidents
“Weeks ago, we launched a thorough investigation into the call’s authenticity,” OpenAI spokesperson Nate Evans said, The Intercept reported.
asserted
Intercept → launch → authenticity
“We acted internally and with the relevant authorities to conclude that it was a prank conducted by individuals with a history of targeting officials and other companies.”
asserted
it → act → officials
OpenAI referred the DCNF to a Sept. 23 announcement that it was giving Ukraine access to tools to protect against cyberattacks.
asserted
it → refer → cyberattacks
The AI company declined to comment on the record when contacted by the DCNF.
asserted
company → decline → DCNF
The Embassy of Ukraine in Washington and the DOW did not respond to the DCNF’s requests for comment.
asserted
Embassy → respond → comment
Vovan and Lexus deny claims that they work for Russia, and said they have faced allegations since their pranks became popular in 2014.
uncertain
pranks → deny → 2014
The Ukrainian Ministry of Defense describes Kuznetsov and Stolyarov as “Kremlin Mouthpieces,” according to Ukraine’s Defence Intelligence (GUR) agency.
uncertain
Ministry → describe → agency
“[M]ore than a decade later, no one has managed to produce any credible evidence to support these claims,” the duo told the DCNF.
uncertain
duo → manage → claims
“Apparently, repeating the same allegations year after year is much easier than actually proving them.”
asserted
repeating → repeat → them
The Aporia
.jpg)
.jpg)
.jpg)
.jpg)