As AI blurs the line between real and synthetic content, trust depends on verification. Here's how a public-private certification model could help consumers understand what they’re seeing.
is the Arthur J. Goldberg Distinguished Professor of Law at UC Law SF and Director of the AI Law & Innovation Institute.When it comes to AI, the name of the game is “verification.” How in the world can we trust what is going on, or that we even know what is going on? Without verification, society cannot develop the trust that is needed to benefit from AI’s potential.
Somehow, we must develop confidence that what we think is real, is real.How can we trust digital content when AI can now generate or manipulate what we see and hear, often with no visible signal that anything has been altered? The avalanche of information can be blindingly blurry. We can look to industry as it develops various forms of trust tech, but scattered approaches are unlikely to bring the widespread assurance that society needs to have confidence in the source and verifiability of content. Even high-quality products may have their brand perception tarnished by results from ostensibly similar competitors using AI of variable quality. Consumers want clarity around where information comes from and how much of it has been generated or modified by AI. We already have models for this kind of assurance, ranging from private trust marks like the Good Housekeeping seal to public-quality oversight from agencies like the FDA.Well, we can’t just ask, "Did they use AI?" AI tools are already in use everywhere, from color balancing to grammar correction. It is no longer a yes or no question. Instead, an effective certification body would establish a series of measures for digital information, so that people have a way to understand the “nutritional quality” of what they’re consuming. Over time, a small standardized box, similar in size and familiarity to the Nutrition Facts label, could become a universally recognized trust marker for evaluating information products. This is not to suggest that the verification body would provide an analysis of whether the contents represent “the truth.” I believe that absolute truth exists; I just don’t trust anyone to tell me what it is. Instead, the verification body would disclose key details about sources and methods, thereby enabling consumers to make their own informed choices within the marketplace of ideas.If the source can’t be identified, the product is already suspicious.We can leave room for imagination and satire while clarifying.We don’t want to ding people for simple sound editing, but we do want viewers to be able to choose whether or not they wish to view AI-generated content, and to what extent. In the case of TikTok or YouTube Shorts, this might be a small box that the user can click to expand with the relevant information.The best verification body would be a public-private collaboration. Neither one would be as effective alone. First, the AI industry is an essential partner. No government agency could ever move with the speed and dexterity of industry. And no one knows more about its own products than industry. Industry alone, however, would not be an ideal choice. We didn’t get nutrition labels because companies got together and said, “Gee, wouldn’t this be a nice idea.” It isn’t in their interests. In addition, fair or not, society is unlikely to trust industry to police itself. The interests simply don’t align. And, of course, the companies would have had trouble sitting down together in the first place. When competitors try to collaborate, antitrust lawyers—not to mention antitrust authorities—normally break out in hives. In short, only the government has the regulatory power to bind industry participants, as well as the greatest power to act in the interests of society as a whole. But government, by itself, lacks the expertise to evaluate and keep abreast of this complex and rapidly evolving field. Together, government and industry could power an efficient and effective certification process. This might function like the National Institute of Standards and Technology , and could even be organized within its purview as an additional NIST Laboratory.And now for the toughest part: Where will we get the funding for this certification body and its process, which will require constant updating to stay ahead of emerging content and the latest AI techniques? The federal budget is seriously constrained, and we should try to avoid over-burdening the AI industry. For both economic and national security reasons, we need to press forward on AI to enhance our position in the international race; hobbling our domestic industry could be counterproductive to that effort. So, how does one take a large sum and make it seem small? The answer is to spread it across as large a group of payers as possible. If 100 companies have to pay $1 million each, that hurts. But if 100 million people pay $1 each, it’s not as bad. The key is to identify the appropriately broad population and the manner of collection so that the impact is minimal in every way. For example, one could establish small fees to be collected by internet service providers or by telecom services that offer access to broadband internet access from each of their customers. These small “trust fees” could have minimal impact on each person while raising the amount needed for certification of virtual content. Verification can pave the way for the public to trust AI, which is necessary both for harnessing the potential of AI and for keeping society from tearing itself apart.
United States Latest News, United States Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Governments endorse greater protections for sharks amid concerns about overfishingGovernments at a wildlife trade conference have adopted greater protections for over 70 species of sharks and rays amid concerns that overfishing is driving
Read more »
Transparency turns into tragedy in ‘Oedipus’ updateSpoiler alert: Oedipus kills his father and marries his mother. But in Robert Icke’s contemporary adaptation of the 2,400-year-old Greek tragedy, the real
Read more »
Marvel Star Josh Brolin Defends ‘His Friend’ Trump: ‘No Greater Genius Than Him in Marketing’Source of breaking news and analysis, insightful commentary and original reporting, curated and written specifically for the new generation of independent and conservative thinkers.
Read more »
Governments endorse support for greater protections for sharks, raysGovernments have adopted greater protections for over 70 species of sharks and rays at an international wildlife trade conference. The measures, approved Friday at the Convention on International Trade in Endangered Species of Wild Fauna and Flora in Uzbekistan, would ban the trade in several shark species and increase regulations on others.
Read more »
Governments endorse greater protections for sharksGovernments adopted greater protections for more than 70 species of sharks and rays at an international wildlife trade convention in Uzbekistan.
Read more »
LGBT Community Center of Greater Cleveland presents superhero art exhibit: A Place in the SunSuperhero art exhibit featuring LGBTQ+ youth recently opened at the LGBT Community Center of Greater Cleveland, which is located at 6705 Detroit Ave. in Lakewood.
Read more »