Regulating AI

A state legislator, Scott Wiener, in California has recently proposed legislation to force companies working with AI to test code/software before releasing any product to market.  This is apropos because the business of business is profit and there is often a tendency to rush product/services to market to increase potential profits; consequences be damned.

I suggest the following as an addition to said legislation; any AI related regulation:

Creating and publishing fictitious content using AI technology in any way, shape or form without a disclaimer identifying the use of said technology can be assumed to have been created and published with malicious intent;

Creating, and publishing, an image  that bears any resemblance to an existing person(s), or write and publish text that uses the content from, or identity of, an existing individual(s) without a disclaimer identifying the work as fiction, and including notifications that consent was/wasn't attained, then the fiction can be considered to have been created and published with malicious intent;

If a publication, of any sort, allows fictitious content of their own, or of others, creation to be made available for public consumption on their site, it may be assumed that said content is shared with malicious intent.


I suggest this, and any other carefully considered, regulation of the use of AI to be necessary, because computer/telecommunications technology is incredibly powerful and omnipresent in the world today.  Said technology need not be intentionally malicious to be the cause of very bad consequences, but the results need be thoroughly reviewed to attain reasonable certainty, one way or the other; corrected, as needed.  It would be better if said adverse results were discovered, by sufficient testing, and averted before software/product has been released out into the world.

The federal government would be the best place for said legislation to occur, but they've been hampered, intentionally (by a Republican majority in Congress in 1995), when the Office of Technology Assessment (OTA) was defunded; effectively eliminated.

No legislative official can be expected to be fluent in everything that legislation must address, so investigative agencies (reporting to any legislative body) are essential.  The Republican legislators knew that when they destroyed the OTA.

The Government Accounting Office has created the Science, Technology Assessment, and Analytics (STAA) to do what the OTA had done.  STAA has recently, in 2024, published informative documents related to AI (GAO-24-106246 and GAO-24-107237).  

Sufficient understanding and regulation of technology, including AI, is both possible and essential.