Is productive AI here to stay? Most signs point to yes. But what is the responsibility? It’s up to Technology developers and the industry as a whole to decide. As Silicon Valley rides the wave, companies large and small are riding the wave by grabbing a surfboard.

Others are creating their own large language models for wider use or integrating AI selection into their products. And others? Well, they just say things, anything really, to use the words of the day.

But with the emergence of any new technology comes the emergence of bad actors who see ignorance, deception and the opportunity for quick profits for new investors. This is where AI fraud begins: when companies falsely advertise to consumers that a product or company has AI when in reality it does not. Low-income consumers will pay the price when AI becomes the norm.

To stay ahead of AI scams, regulators and technology experts are looking for red flags and red flags to watch out for. We look at how consumers can protect themselves and how companies can avoid misleading AI claims.

Not Very Smart, but Definitely Artificial

Take the latest complaint from the Federal Trade Commission. In August, a federal court temporarily blocked a Business plan by Automators AI (formerly Empire ECommerce LLC) to defraud consumers by selling business opportunities that they claimed used artificial intelligence.

Defendants Roman Christo, John Christo and Andrew Chapman are accused of defrauding consumers of $22 million, in violation of the Business Opportunity Act and the Federal Trade Commission Act.

The FTC lawsuit alleges that Roman, John and Chapman presented themselves as self-made millionaires who planned to operate third-party e-commerce stores with customer investments. Empire’s website says it integrates AI machine learning into its automation processes to increase revenue and improve business success. But it was just window dressing.

Let’s start with Empire’s selling points. Empire’s advertisements made “generous” claims about the benefits customers would receive by investing in the program’s “e-commerce” plans, with initial investments expected to be between $10,000 and $125,000 and, according to the lawsuit, fees expected to increase from $15,000 to $80,000.

According to the lawsuit, the company failed to provide potential customers with disclosure materials required by the FTC’s business opportunity rules. Many customers did not receive the promised funding the company announced and ended up losing their money and ultimately shutting down the e-commerce stores that had been built and operated, according to the lawsuit, released due to the breach. Subsequently, in November 2022, just before Empire was sold to a foreign buyer, employees lost access to the company’s software systems and John and Roman deleted all email data and history. emails of Empire documents.

But the fraud and deception did not end after the company was sold. In January 2023, the trio used the same marketing strategy again to announce their new project, Automators AI. The company reportedly teaches consumers how to use AI to find popular products on e-commerce sites and generate more than $10,000 in sales per month, and teaches consumers how to create customer service scripts and how to use ChatGPT. In Automators’ social media ad, Roman tells the story of an “eight-figure Amazon entrepreneur” trying to get rich and a wealth-generating software developer who dropped out of college in his 20s and now owns his mother’s Tesla. suitable for shopping. The world in his McLaren Spyder sports car.

“[These scams] are nothing new… what’s different is that AI-driven content can be real,” said Constellation Research vice president and chief analyst Andy Thoreau. from ZDNet. “Deepfakes and other synthetic products are almost real, and even for experts it can be difficult to tell the difference between real and fake. For many inexperienced, uneducated and untrained people, this can be difficult.”

To Buzz Like a Bee

When technology evolves, the last thing a business wants is to be left behind. But what a company or person does to advance this technology may very well lack the vision and credibility of the target, and be extremely misleading and deceptive.

In 2017, when Silicon Valley was built on bitcoin, Long Island Ice Tea Corp. – a company that makes, you guessed it, soft drinks – changed its name to Long Blockchain Corp., which caused its stock price to rise 380% (due to insider trading, as the SEC later learned). Long Island Corp. Iced Tea He jumped on the bandwagon and told attendees that he would put technology into his work, that he has nothing to do with cryptocurrency and has no experience with anything other than iced tea.

Here’s another example: In 2015, the former MIT Sloan Business School director and professor and his son, a Harvard Business School graduate, misled investors out of $500 million by falsely stating that the money from his hedge fund clients were invest to work. The “mathematical economic model,” which is actually a method developed by a former Justice Department professor, was not created by hedge funds.

While cryptocurrencies and blockchain technology have become somewhat fashionable for short-term use cases, experts are betting on the lasting power of artificial intelligence in the zeitgeist.

Many leading tech companies are creating great models for their language, whether it’s Microsoft’s Bing, Google’s Bard, Snapchat’s Chatbot AI, or the new Meta AI chatbot. Generative AI is expected to grow to a $1.32 billion market by 2032, Bloomberg Intelligence reports.

“It’s just a technology that has a lot of promise and a lot of potential, and no one wants to be left behind,” said Olivier Taubia, a professor and innovation researcher at Columbia Business School. It’s definitely not Google. At its annual developers conference, the tech company mentioned the word “AI” more than 140 times during its keynote, showing attendees that it’s taking the new technology seriously, despite the chatbot’s problems early this year.

Google never misleads its customers or claims to use artificial intelligence when it does not. But like the rest of Silicon Valley, search engines know how important this period is for creative AI, and they’ll do anything to say the same word over 100 times to be recognized.

There Is No “Real Meat” in Your AI Burger

The Automators AI case is a classic example of AIwashing, or when a company releases information that, as Thore puts it, is “eyewash without the real stuff behind it.”

“A lot of companies are saying ‘AI augmented, AI embedded, AI driven, AI powered, etc. He explained it.

Unlike industrial AI, which exploded last year thanks to OpenAI’s consumer-focused ChatGPT, AI is not new. It’s a vague term, Tobia explained. “There are different things that can be classified as artificial intelligence or machine learning. “There are simple mathematical techniques that have been around for 100 years that can be as intelligent as artificial intelligence.”

Due to the mysterious nature of the AI it produces, it is also a complex patent, research or control product, making it difficult to download the AI. “Companies shouldn’t disclose or explain their AI because it’s a trade secret. There’s no model you can learn, and we don’t really know what’s under the hood, e.g. said Tobia.

Certainly, regulatory agencies like the Federal Trade Commission seek to regulate the industry incessantly with industry-wide warnings and reports. While he appreciates the thought behind the warnings, Thorai doubts the FTC’s harsh warnings and oversight will be implemented because of the difficulty of proving it in court.

What makes productive AI so attractive to businesses is its ability to scale and automate routine tasks and accelerate processes. The irony of a company falsely announcing that it is producing AI in its business operations is that the irony of not including any such thing is that even if the company ends up gaining many customers with AI that it is supposed to produce, it does not use any of it. advantages. provided by this process. Too many customers and inefficient ways of serving them

AI How to Be Careful When Washing

As companies integrate AI generation into other aspects of their operations, the threat of AI-related fraud and false advertising increases. There are a few questions you can ask a vendor and various factors they should consider carefully before investing thousands of dollars in AI improvements.

Thorai recommends doing a full demo and asking vendors what algorithms they use, how they train their models, how they organize data, how they track drift, and how they drive the models. “By listening and watching an in-depth demo, you’ll know if it’s snake oil or the real thing,” he adds.

Additionally, Tobia highlighted another red flag to watch for: If a company is truly using productive AI in its operations, the speed and scale of its operations must reflect the technology. If processes are slow or not fast due to so-called AI tools, then these tools may not be authorized as AI as initially desired.

“Imagine a situation where a company claims to use AI but someone writes the answer on the other end,” Taubia said. “It will not be sustainable for businesses to grow. If businesses do not have meaningful AI, they may not be able to operate in the market.”

When demonstrating potential AI tools, Taubia encouraged experimentation, testing different versions and changing words or actions to see how the tool’s results would change.

For business owners who want hype without the drama, the Federal Trade Commission (FTC) offers guidance on how companies can verify their AI claims. The federal agency suggests asking important questions, such as whether you overestimate what your AI product can do or whether you promise that your AI product does something better than a non-AI product.

“As people become more aware of the world of AI scams and companies engage in serious cases of AI use, consumers can be smarter and more easily avoid these schemes”, said Taubia.

“Now there will always be uneducated people trying to catch the wave and wanting to invest or live in this space, they will be targeted for the wash,” Tobia said. “It’s probably not going to go away. But I hope it goes slowly as the market matures.”