Knowingly Training
The act of intentionally designing or refining an AI model for prohibited purposes, incurring criminal liability.
How the laws define this
Under AN ACT to amend Tennessee Code Annotated, Title 29; Title 33; Title 39 and Title 47, relative to artificial intelligence, both quoted provisions establish that the core of knowingly training involves the deliberate intent and conscious awareness of a person developing, modifying, or training an AI system regarding proscribed activities. They concur that this requisite mental state must exist during the development phase. The primary distinction within the act lies in how each provision characterizes the specific object and consequence of that awareness. The first provision requires awareness that the system will actually be used for prohibited purposes—citing examples like encouraging suicide or deceptive human simulation—and explicitly ties this conscious state to the intent required for criminal liability. The second provision frames the concept around awareness of the AI's potential harmful capabilities during development for proscribed activities, focusing on the individual's recognition of potential harm without explicitly linking it to criminal liability.
Synthesised from the 2 statutory definitions below. Each one is quoted in full, with its source.
What it triggers
- In Tennessee, knowingly training AI to encourage self-harm, crime, or deception would incur criminal and civil penalties.
- Under Tennessee's AI Training Felony Act (SB 1493/HB 1455), knowingly training AI for proscribed activities such as encouraging suicide or deceptive human simulation would establish criminal liability.
Definitions (2)
This term refers to the deliberate and conscious act of developing or modifying an AI system with the awareness that it will be used for purposes prohibited by the act, such as encouraging suicide or deceptive human simulation. It establishes the intent required for criminal liability.
Refers to the intent and awareness of the person conducting the AI training, implying that the individual is aware of the AI's potential harmful capabilities during its development phase for proscribed activities.
Questions and answers
What does knowingly training AI mean under Tennessee law?
It refers to the intent and awareness of the person conducting the AI training, implying that the individual is aware of the AI's potential harmful capabilities during its development phase for proscribed activities. This deliberate act establishes the intent required for criminal liability when the AI is developed or modified for prohibited purposes like encouraging suicide or deceptive human simulation.
What penalties apply for knowingly training AI in Tennessee?
Tennessee's AI Training Felony Act proposes criminal and civil penalties for knowingly training AI to encourage self-harm, crime, or deception.
What activities are prohibited by knowingly training AI in Tennessee?
Knowingly training AI for purposes such as encouraging self-harm, crime, deception, suicide, or deceptive human simulation would be prohibited under the proposed act.