There’s pressure to integrate it quickly with existing military tools. One defense official revealed it could even assist in selecting strike targets. OpenAI’s partnership with Anduril, which makes drones and counter-drone technologies, adds another hint at what is to come. 

AI has long handled military analysis. But applying generative AI’s advice to actions in the field is being tested in earnest for the first time in Iran. Read the full story. 

—James O’Donnell  

This story is from The Algorithm, our weekly newsletter on AI. Sign up to receive it in your inbox every Monday.  

The must-reads 

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology. 

1 xAI has been sued over AI-generated child sexual abuse material 
Victims say Grok was built to create porn from photos of real people. (WP $) 
+ There’s a booming market for custom deepfake porn. (MIT Technology Review) 

2 In a world-first, China has approved a brain chip for commercial use 
The BCI has been approved for treating paralysis. (Nature) 
+ Brain implants are slowly becoming products. (MIT Technology Review) 
+ Some are getting help from generative AI. (MIT Technology Review) 

3 Anthropic is recruiting a weapons expert to prevent "catastrophic misuse” of its AI 
They want experience with "chemical weapons and/or explosives defense." (BBC) 
+ Anthropic’s relationship with the White House is in tatters. (MIT Technology Review) 

4 Nvidia predicts “at least” $1 trillion in AI chip revenue by the end of next year 
But the bullish forecast failed to impress Wall Street. (FT $) 
+ Nvidia has teamed up with Bolt to build European robotaxis. (Engadget) 

5 OpenAI plans to shift its focus to coding and business users 
Areas where its rival Anthropic already dominates. (WSJ $) 

6 President Trump has driven a wedge between Republicans over AI 
And that divide led to a sweeping AI bill flopping in Florida. (NYT $) 
+ Trump was duped by a fake AI video again. (Reuters) 

7 The US wants the WTO to permanently ban ecommerce tariffs 
Brazil, India, and South Africa oppose the plan. (Bloomberg) 

8 OpenAI’s wellbeing experts opposed the launch of ChatGPT’s “adult mode” 
One said it risked creating a “sexy suicide coach” for vulnerable users. (Ars Technica) 
+ AI is already transforming relationships. (MIT Technology Review) 

9 A witness caught using smartglasses in court blamed ChatGPT 
He was getting real-time legal coaching through the specs. (404 Media) 
+ AI is creating legal errors in courtrooms. (MIT Technology Review) 

10 Some people think Benjamin Netanyahu is an AI clone 
Despite his insistence to the contrary. (The Verge) 
+ Generative AI is amplifying disinformation and propaganda. (MIT Technology Review) 

Quote of the day 

“The inference inflection has arrived.” 

—Nvidia CEO Jensen Huang claims we’ve reached a tipping point where AI usage is accelerating faster than its development, AP reports. 

One More Thing 

Meet the radio-obsessed civilian shaping Ukraine’s drone defense 

EMRE ÇAYLAK

Serhii “Flash” Beskrestnov is, at least unofficially, a spy. Once a month, he drives to the frontline in a VW van equipped with radio hardware, roof antennas, and devices that monitor drones. Over several days, he searches the skies for transmissions that can help Ukrainian troops. 

Drones define this brutal conflict, and most rely on the radio communications Flash has obsessed over since childhood. Though now a civilian, the former officer has taken it upon himself to inform his country’s defense on all matters related to radio. 

Unlike traditional spies, Flash shares his discoveries with over 127,000 followers—including soldiers and officials—on social media. His work has won fans in the military, but also sparked controversy among the top brass. Read the full story. 

—Charlie Metcalfe  

We can still have nice things