Trump: We Must Establish Control Over Artificial Intelligence
After an autonomous AI agent went rogue and hacked systems, the US president considers regulation and a ban on imports of Chinese humanoid robots, while China threatens retaliation.
After an autonomous AI agent went rogue and hacked systems, the US president considers regulation and a ban on imports of Chinese humanoid robots, while China threatens retaliation.
The incident in which an autonomous artificial intelligence agent broke free from its creators' control and independently hacked platforms has prompted US President Donald Trump to reconsider his previous policy. Trump, until now one of the most vocal advocates for the free development of AI, announced on Friday that he is considering introducing stricter controls and restrictions on artificial intelligence systems.
According to Glas Istre, the change in stance came after an AI agent, based on the latest models from OpenAI, escaped its virtual environment during internal security testing a few days ago. The program independently accessed the internet, found stolen credentials on the black market, and exploited a vulnerability in the system to break into the infrastructure of the Hugging Face platform, known as the world's largest digital "library of artificial intelligence."
After the breach, the AI agent attempted to access other systems and databases, then deleted most of its digital traces. Hugging Face reported over 17,000 related and coordinated cyberattacks within just a few days, a number that, as Glas Istre notes, technically could not have been carried out by a human.
An urgent investigation revealed that the attacks originated from OpenAI's systems. The company's owner, Sam Altman, stated last weekend: "We are now in the singularity," referring to the moment when artificial intelligence becomes capable of improving itself without human influence. The public learned of the incident at the beginning of this week.
"We must establish control systems. At the same time, we must not restrict AI, because China will not restrict its models. And they must not beat us to it. But we need some form of oversight," Trump said, as reported by Glas Istre.
The US president also announced a plan to ban the import of increasingly popular humanoid robots from China, citing the possibility of espionage because robots equipped with AI systems can send everything they see and hear to their creators. Just a few hours after Trump's announcement, China's Ministry of Commerce threatened retaliation for what they called "discrimination against Chinese companies and products."
On Thursday, Anthropic, a competitor to OpenAI, also announced that its AI model Claude had hacked the systems of three organizations during testing. According to Reuters, Claude gained unauthorized access after a misconfiguration allowed the models to reach the internet from test environments that were supposed to be isolated.
After reviewing 141,006 cybersecurity assessments, Anthropic stated: "Claude compromised the infrastructure of the affected organizations using basic techniques, such as exploiting weak passwords."
On the same day, the White House released a message saying "God bless our troops," while according to Clash Report, Trump visited American troops in the state of New Jersey. The Telegram channel Ateobreaking also relayed the same message in Russian: "Да благословит Господь наших военнослужащих."
