Issued on 09/03/2026 - FRANCE24
PLAY 06:43 min
A struggle to control artificial intelligence is playing out just as the United States increasingly deploys the technology in conflicts from Venezuela to Iran.
In its interventions in Venezuela and Iran, the United States military has reportedly used Anthropic's Claude chatbot to analyse battlefield data. At the same time, a monumental argument has broken out over the future capabilities of the game-changing technology.
Donald Trump's administration has ripped up its partnership with Anthropic and banned defence contractors from using Claude, after Anthropic insisted it should not be used for fully autonomous weapons or for mass surveillance of US citizens. These are not current uses, but the Pentagon notably sees ruling them out as an obstacle to keeping up with China.
For Samuel Hammond, Chief Economist at the Foundation for American Innovation, the ban is counterproductive.
"Designating Anthropic a supply chain risk is usually reserved for adversaries," he said, "for Chinese companies seeking to attack our systems".
It's the first time such a sanction has ever been applied to an American company. And now, the Trump administration has drafted strict rules for AI companies to obey in future government contracts, according to the Financial Times.
Hammond argues that Anthropic's market success contradicts the depiction of the company as radical "leftwing nut jobs" by President Trump.
"There's misinformation going on within the Department of Defense, within elements of the White House that believe Anthropic is a very left-wing company with extreme views," said Hammond, "When, in fact, they are currently the number one download on the App Store."
Anthropic has threatened to sue the Pentagon, while also trying to contain the fallout and defend its values.
When asked whether right-wing libertarian think tanks such as his had helped boost support for Trump, paving the way for him to wield power in this way, Hammond said, "This feels like a betrayal, both of their stated mission [in AI policy] and of broad libertarian values. It's not a libertarian value to seek to destroy a particular company."
"Who is the ultimate arbiter of how these tools are used? ... This is a longer-term process that we're going to have to work out through the democratic process."
Watch this week's Tech 24 for more on what happens when a power-hungry government is confronted with the possibility that someone else might want to call the shots on the most useful and promising technology of the age.
BY: Peter O'Brien
No comments:
Post a Comment