Direct Translation via Google Translate. Edited
[Regnum] The heads of major tech companies have agreed with US President Donald Trump to introduce a four-tiered system of oversight for the safety of advanced artificial intelligence models. The White House chief announced this on the social media platform Truth Social.
The agreement was signed by Google CEO Sundar Pinchai, Anthropic CEO Dario Amadei, Meta CEO (recognized as extremist and banned in Russia) Mark Zuckerberg, OpenAI President Greg Brockman, xAI CEO Elon Musk, and Nvidia CEO Jensen Huang.
"Every company training and deploying advanced models must have robust internal processes and controls in place," the agreement states.
The companies have committed to creating their own mechanisms to monitor the capabilities of AI models and their compliance with security requirements. Particular attention will be paid to cybersecurity risks, biological and chemical threats, and preventing unintended access of models to technical systems.
Independent committees of boards of directors will provide another level of oversight. They will receive the results of internal and external audits and oversee the remediation of any identified deficiencies.
The parties to the agreement also intend to meet regularly, develop common standards, and exchange practices in the field of artificial intelligence security.
Musk has previously warned of the risks associated with the uncontrolled development of AI. According to him, the uncontrolled development of neural networks could lead to the destruction of humanity. The businessman compared the proposed future scenario to the plot of the film "Terminator."
# id) ?>
|
