In order to verify that AI systems generate reliable results, the Biden administration is now reportedly exploring legislation that would mandate government audits.
These audits might include evaluations of whether AI is encouraging “misinformation” and “disinformation,” among other things. Words that many now believe may be simply code for the Biden administration’s woke ideology/
Government audits of AI systems are one way to increase trust in this new technology, according to Alan Davidson, assistant secretary of communications and information at the National Telecommunications and Information Administration (NTIA) of the Commerce Department.
“Much as financial audits create trust in the accuracy of financial statements, accountability mechanisms for AI can help assure that an AI system is trustworthy. Policy was necessary to make that happen in the finance sector, and it may be necessary for AI.” Davidson said.
Davidson made the statement during a speech this week at the University of Pittsburgh.
Federal authorities will undoubtedly face both technological and political obstacles in their efforts to control the misinformation produced by AI.
Government authorities are only starting to think about how to govern AI; for instance, the NTIA just announced this week that it was looking for public feedback on how it should go about doing so.
Furthermore, since Republicans and Democrats can’t seem to agree on a definition, the topic of misinformation and disinformation has become quite contentious under the Biden administration.
Even though the effort may be challenging, concerns about biased AI have grown over the past year as the technology has developed and may be ripe for government regulation.
Elon Musk, the CEO of Tesla and SpaceX, expressed his concern last year that AI employed in ChatGPT systems was being developed to filter out any output that would offend anyone.
Musk claimed that the Biden administration’s moves presented the risk the coming rise of a “woke” AI.
A voluntary risk management framework provided by the National Institute of Standards and Technology and an AI Bill of Rights are two examples of the preliminary recommendations the Biden administration has released so far on how AI should be developed.