Podcast - The FTC Aims for True Fairness in Use of AI

Good day and welcome to another podcast of Clearly Conspicuous. As we've noted in previous sessions, our goals in this podcast is to make you succeed in this environment, make you aware and knowledgeable about what's going on and give you practical tips for success. As always, it's a privilege to be with you today.

The FTC's Approach to AI and the Tools It Uses for Enforcement

Today we continue our discussion of artificial intelligence or AI Today's topic is what the Federal Trade Commission is saying about it. Essentially, the FTC has focused on the issue of AI generally and the claims being made about AI. So let's start with what the FTC calls aiming for truth, fairness and equity in a company's use of AI. About a couple of years ago, the FTC issued a bulletin where they said the following: "Advances in artificial intelligence technology promised to revolutionize our approach to medicine, finance, business operations, media and more. But research has heightened how apparently neutral technology can produce troubling outcomes, including discrimination by race or other legally protected classes." So then the FTC goes on to say, "The question then is how can we harness the benefits of AI without inadvertently introducing bias or other unfair outcomes?" And the FTC then instructs that it has decades of experience enforcing three laws that are important to the developers and users of AI. There's Section 5 of the FTC Act that prohibits unfair or deceptive practices, and that would include the sale or use, for example, of racially biased algorithms. The FTC also enforces the Fair Credit Reporting Act, and the FCRA comes into play in certain circumstances where an algorithm is used to deny people employment, housing, credit, insurance or other markets. The FTC also enforces the Equal Credit Opportunity Act, and the ECOA makes it illegal for a company to use a biased algorithm that results in credit discrimination on the basis of race, color, religion, national origin, sex, marital status, age or because a person receives public assistance.

The Legal Implications of AI

So the FTC wants to set out certain parameters that we need to think through about the legal implications of AI. And they say, well, let's start with the right foundation. If a data set is missing information from particular populations, using that data to build an AI model may yield results that are unfair or inequitable to legally protected groups. Next, the FTC warned to watch out for discriminatory outcomes. In addition, the FTC says, embrace transparency and independence. And the FTC says as your company develops and uses AI, think about ways to embrace transparency and independence, for example, by using transparency frameworks and independent standards, by conducting and publishing the results of independent audits and by opening your data or source code to outside inspection. Or the FTC says don't exaggerate what your algorithm can do or whether it can deliver fair or unbiased results. For example, the FTC says, let's say that an AI developer tells clients that its products will provide 100 percent unbiased hiring decision, but the algorithm was built with data that lacked racial or gender diversity. The result, therefore, may be deception, discrimination and an FTC law enforcement action. And finally, the FTC says, tell the truth about how you use data. The FTC has given guidance that advise businesses to be careful about how they get the data that powers their model. The FTC has noticed a complaint against Facebook, for example, which alleged that the social media giant misled consumers by telling them that they could opt in to the company's facial recognition algorithm when, in fact, Facebook was using their photos by default.

Deceptive AI Marketing Claims

So now let's talk about the marketing claims of AI. About a month ago, the FTC issued a bulletin saying, keep your AI claims in check. The FTC has said that it is often referring to a variety of technological tools and techniques that use computation to perform tasks such as prediction decisions or recommendations. But the FTC said one thing's for sure: AI is a marketing term. So AI hype is playing out today across many products. The FTC notes from toys to cars to Jackbox and to a lot of other things in between. And that in some products, AI claims that it might even work as advertised in the first place. So the FTC is warning that when you talk about AI in your advertising, the FTC may be wondering, among other things, are you exaggerating what your AI product can do? The FTC notes that your performance claims would be deceptive if they lacked scientific support, substantiation or if they apply only to certain types of users or under certain conditions. The FTC then asks, are you promising that your AI product does something better than a non-AI product? The FTC notes that you need adequate proof for that kind of comparative claim and that if such proof is impossible to get, don't make the claim. In addition, the FTC asks, are you aware of the risks? The FTC notes that companies need to know about the necessarily foreseeable risks and the impact of your AI product before putting it out on the market. If something goes wrong — maybe it fails or yields biased results — you can't just blame a third party developer or the technology, and you can't say you're not responsible because that technology is a black box you can't understand or didn't know how to test.

A Product Using AI vs. a Product Developed by AI

Finally, the FTC asks, does the product actually use AI at all? And the FTC warns before labeling your product as AI-powered, know that merely using an AI tool in the development process is not the same as a product having AI.

Conclusion