The Nationwide Institute of Requirements and Know-how (NIST) has introduced a brand new pilot analysis program, NIST GenAI, to assist assess whether or not content material — textual content, picture, video, or audio — was generated by a human or AI.
One of many objectives of the research is to make use of the outcomes to help folks in making these determinations of their day-to-day lives.
The outcomes of the research is also used to create content material authenticity detection applied sciences and applied sciences that may determine the supply of faux data.
NIST GenAI will contain Generator groups and Discriminator groups. The Generator groups will likely be evaluated on their system’s means to make use of AI to generate content material that may’t be differentiated from human-generated content material. The Discriminator groups will likely be evaluated on their system’s means to detect AI-generated content material.
NIST’s GenAI is a part of the response to President Biden’s Government Order on AI and the outcomes will inform the work of NIST’s U.S. AI Security Institute.
“Pilot evaluations present helpful classes for future analysis on cutting-edge applied sciences and steerage for accountable and secure use of digital content material,” the NIST GenAI web site states.
