You have undoubtedly heard of the Industrial Revolution, the first of the more recent technological revolutions. When we started to write this article, we did a bit of research about the technological revolutions that have occurred in the last 300 years as a lead into this column. To our surprise, we learned that some writers say we have had three and now find ourselves in the fourth, while others say we find ourselves in the sixth. Exploring those differences and the full history of technology revolutions exceed the scope of this article.
If you prefer to think in terms of “Ages” rather than “Revolutions,” think of AI as the latest evolution of the Information Age, which has converted information into a commodity capable of rapid and wide distribution, usually using computers and related technology.
Whether you like to think in terms of “Ages” or “Revolutions,” the increasingly rapid development of artificial intelligence has created the latest. AI has caused many changes in the world, and it will continue to impact us in many ways and have a ripple effect that will reach virtually every aspect of our personal and professional lives.
In previous columns, we talked about the duality of technology in general and AI in particular. By “duality,” we refer to the fact that technology and AI are present as a double-edged sword. Cutting one way can benefit us in untold ways; cutting the other can cause damage ranging from minor to catastrophic in its impact. These benefits and detriments exist at all levels, including personal, professional, business, and, ultimately, societal. In this column, we will focus on AI and its ability to fuel or support scams, particularly since the release of ChatGPT and the expansion of generative AI (GenAI).
AI currently comes without a moral compass and does the work assigned to it by humans. Unfortunately, humanity includes many bad actors. Many have acquired technological proficiency that enables them to capitalize on the AI’s functionality in ways that facilitate taking advantage of others using one of many techniques we refer to collectively as “Scams.”
A recent article in the San Francisco Chronicle (February 11, 2024, Section A, Page 7) discusses one example of this problem. That article, “Feds Targeting AI Scam Calls” by Chase DiFeliciantonio, identifies robocalls using AI technology to generate voices as the number one consumer complaint. The calls sometimes sound like a famous or familiar person. The technology uses both audio and text messaging. A Truecaller/Harris Poll cited in the article found that “[t]he number of phone scam victims in the U.S. nearly tripled between 2017 and 2022.” The article also notes that the FCC revealed that in 2022, fraudulent calls and texts generated $1.13 billion in known losses.
The bad guys have used the technology to fool voters, to try to collect utility bills falsely claimed as “unpaid,” to imitate the voices of other family members and friends in pleas for financial assistance for innumerable reasons, and a myriad of other schemes. This scam, often called the “grandparents’ scam” because it has proven successful when used to contact grandparents pretending to be a troubled grandchild, preys, particularly on the elderly, trying to convince grandparents that a grandchild needs their help for a variety of reasons running the gamut from transportation issues to needing bail money. It does not limit itself to grandparents and the elderly, however, and one does not need to have reached a minimum age to acquire vulnerability or to become a target.
Respecting billing issues, we have seen calls purporting to come from utility companies, insurance providers, and government agencies respecting allegedly unpaid bills.
The Senior Lawyers Division has identified scams as a major concern and will emphasize that concern in its activities during the current bar year. It recently approved a book on scams that should be available early next calendar year. In the interest of full disclosure, we will edit that book, which will be a compilation of contributions from several authors. We will also contribute to it as authors.
In the bad old days, most of these scams existed but ran using a human making a call and trying to sell themselves as whatever the fraud required. Some people have an acquired ability to mimic the voices of others. Many entertainers support themselves by imitating famous people. As good as some have been, AI does it much better. AI can produce an almost perfect replication of the voice of another. Unlike humans, who require time and practice to mimic a voice well, AI can take a few voice samples and instantly create its version. Some of these scams can and do still work without AI. For example, it does not take an AI voice imitation to present as a representative of a government agency or a utility company. They can sound like anybody and make the play. Selling oneself as a close friend or relative poses a far more difficult proposition as the victim will be familiar with this voice. However, the availability and evolution of AI made those scams more common by creating the ability to use an AI representation to make many calls in a short time frame, far more than one or a handful of people could make on their own.
Some scams would never have worked well before the recent evolution of AI. For example, a recent scam used an excellent AI-generated imitation of President Biden’s voice to call voters and discourage them from voting in the New Hampshire Primary.
The voter scam tipped the balance, and now, the Attorneys General of 26 states have asked the FCC to clamp down on this abuse. The FCC responded by outlawing robocalls using AI-generated voices. We support the FCC’s action, but we do so, tongue in cheek. While this seems like a step in the right direction, outlawing AI-generated robocalls does not mean they will stop. It only means the government can prosecute those who do it and get caught.
Efforts to stop or limit robocalls, scam telephone calls, and even calls by live people trying to sell you a product or service that may or may not be legitimate have generally and regularly failed. How many of you used the “Do Not Call” system to prevent calls from telemarketers? How many of you stopped getting them? You can get software to help you reduce such calls on your mobile phone, but the technique has not stopped them. The situation has reached the point that we get multiple calls of this nature every day. We have reached the point that we will not answer the phone unless the caller has a caller ID and is a person or is calling from a number we recognize. If we miss a call we need to get, the caller will likely leave a message, and we can call back, if appropriate.
Given that you will likely continue to get such calls and texts, you should protect yourself as much as possible. While you cannot stop the calls from coming, you can act to reduce your exposure to them and your susceptibility. We do not intend the following as a comprehensive list of everything you can do to protect yourself. An exhaustive list would require more space than we have available. We do, however, recommend that, at a minimum, you follow these suggestions and remember that avoiding scams always involves vigilance, work, and a healthy dose of skepticism and that the use of AI, particularly GenAI, makes that even more imperative:
