New Wisdom Times Report
Editor: Aeneas is very tired
[New Wisdom Times Overview] Recently, a bill passed by the U.S. House of Representatives has imposed stricter restrictions on individuals working in AI in the U.S. If the bill is passed, Chinese individuals working in AI in the U.S. may be affected.
Recently, a bill in the U.S. House of Representatives has sparked great controversy.
This "Enhancing Foreign Nations' Understanding of Export Controls (ENFORCE) Act," as speculated by netizens, may impact a large number of Chinese individuals working in AI in the United States.
ENFORCE Act Details
The purpose of introducing this act is to prevent foreign adversaries from using American AI or other technologies for alternative purposes. Definition
The act proposes to include the definitions of artificial intelligence, AI systems, and related AI systems into the 2018 Export Control Reform Act.
It defines an AI system as, "‘Artificial intelligence system’ means any software or hardware that implements artificial intelligence, including artificial intelligence model weights and any numerical parameters associated with the implementation of artificial intelligence."
The interim definition in the act states that a related AI system is one that exhibits or could be modified to exhibit high-level capabilities to perform tasks that pose a significant risk to U.S. national security and foreign policy or a combination of these issues, even with technical safeguards in place when provided to end-users.
These capabilities include:
The act specifies that in finalizing the definition of "related AI systems," the Secretary should consider technical and non-technical factors in consultation with the Secretary of State, Secretary of Defense, and Secretary of Energy to most effectively promote U.S. national security.
Power Granted to the President
The act grants the President the authority to "control the activities of U.S. persons anywhere engaging in activities related to specific related AI systems and emerging and foundational technologies deemed critical to U.S. national security under section 1758(a) of the Export Control Act."
In essence, the U.S. President has the power to control the activities of U.S. citizens, regardless of their location, as long as they involve specific AI systems and technologies deemed crucial to U.S. national security.
Licensing of Exporting AI Models
According to the act, U.S. individuals must apply for a license from the Department of Commerce, regardless of their location, to export, re-export, or transfer related AI systems and specific emerging and foundational technologies critical to U.S. national security. The Bureau of Industry and Security within the Department of Commerce will play a crucial role in granting export licenses.
The Department of Commerce must also license activities that may support the "design, development, production, use, operation, installation, maintenance, repair, overhaul, refurbishing, or provision of services" related to AI models.
According to speculation by a user on the "One Point Three Acres" forum, if the act passes, Chinese individuals holding H-1B visas may require special permits to work in AI/ML-related fields.
Previously, hardware engineers in the chip/semiconductor industry used to be required to apply for an "export control license" before starting their jobs.
The "ENFORCE Act" this time has expanded the scope of control that was previously in place.
The U.S. "AI Ban"
Previously, the United States has implemented a series of measures to mitigate the national security risks posed by AI.
In October 2023, President Biden issued an executive order that stipulated developers must share their security testing results and other critical information with the U.S. government under the Defense Production Act before developing "the most powerful" AI systems.
The Department of Homeland Security (DHS) in the United States has even established a framework for nucleic acid synthesis screening to prevent the misuse of AI in developing harmful chemical and biological materials.
This executive order emphasizes that AI systems could pose threats to critical infrastructure and exacerbate risks in chemical, biological, radiological, nuclear, and cyber security.
In early May, nine government agencies announced the initiation of AI risk assessments for 16 critical infrastructure sectors and developed AI security guidelines for owners and operators of critical infrastructure.
On April 26, the Department of Homeland Security prominently established an AI Security Committee, with members including CEOs from major AI companies such as OpenAI, Anthropic, NVIDIA, Microsoft, Google, Amazon, and AMD.
The committee is dedicated to providing recommendations to various departments to prevent and prepare for potential AI disruptions that could impact national security, economic stability, public health, and safety.
