“It will solely embolden different militaries or adversaries to do the identical.”
Pentagon Friends
Pentagon-operated AI personas is perhaps coming to an web discussion board close to you.
As The Intercept reports, a wishlist put ahead by the Joint Particular Operations Command (JSOC) — a furtive counterterrorism group inside the US Division of Protection (DoD) — reveals the company’s curiosity in utilizing generative AI to create faux on-line web customers.
That is regardless of the US authorities’s persistent warnings that deepfakes and different AI-generated content material will deepen the misinformation disaster and result in a muddier info ecosystem for everybody.
Within the doc, JSOC explains that it is looking for “applied sciences that may generate convincing on-line personas to be used on social media platforms, social networking websites, and different on-line content material” to be used by Particular Operations Forces. This “answer,” JSOC provides, “ought to embrace facial and background imagery, facial and background video, and audio layers.”
In response to the wishlist, JSOC desires SOF brokers to “use this functionality to assemble info from public on-line boards.”
In different phrases, JSOC desires to supply its particular ops groups with the expertise to create refined, bespoke AI deepfakes, able to convincing social media customers that they are the true deal for the sake of information-gathering efforts.
AI Influenza
On one hand, it isn’t like intelligence personnel do not lurk round on-line message boards and social media channels already. The Pentagon can be no stranger to external online influence campaigns.
However the DoD’s continued concentrate on AI as a option to degree up its digital surveillance and affect efforts is a noteworthy improvement in its strategy to on-line intelligence operations.
As The Intercept reported last year, the Pentagon expressed curiosity in deepfakes as a way of enhancing and increasing affect efforts run by the DoD’s Particular Operations Command (SOCOM), writing in a 2023 procurement that it sought “extra encompassing, disruptive” applied sciences “bigger in scope” than then-current instruments.
And with election day swiftly approaching, it is value noting that Mission 2025 — a blueprint for a projected Donald Trump presidency penned by dozens of close allies — salivates over the prospect of utilizing AI to develop surveillance and spying efforts, as a Futurism deep-dive into the coverage playbook confirmed.
However because the Pentagon races to make use of AI to supercharge its murky digital actions, consultants are warning that the American protection sector’s embrace of the tech will greenlight related misleading practices for nations across the globe.
“It will solely embolden different militaries or adversaries to do the identical,” Heidy Khlaaf, chief AI scientist on the AI Now Institute, informed The Intercept, “resulting in a society the place it’s more and more tough to establish fact from fiction and muddling the geopolitical sphere.”
Extra on AI and surveillance: Billionaire Drools That “Citizens Will Be on Their Best Behavior” Under Constant AI Surveillance
