Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Raising the alarm: Artificial intelligence has a digital identity problem

As AI advances, we must recognize the challenges posed by the uncanny valley and the risks that could fuel the next generation of digital identity fraud.

[Images Adobe Stock/Marcela Ruty Romero]

Profile photo of Rodger Desai

Have you ever caught a glimpse of a robot so life-like that it gives you the creeps? If so, you’ve fallen into the “uncanny valley,” according to Professor Masahiro Mori of the Tokyo Institute of Technology. In his research, Mori observed that humans balk at robots that are too lifelike, describing them as eerie. Although the “uncanny valley” was coined to describe the uneasy relationship between humans and life-like robots, I would argue that it also perfectly encapsulates our current relationship with artificial intelligence (AI).

With AI shooting into the mainstream, it’s clear we are beginning a prolonged period in the uncanny valley that will transform our digital world. To ensure that the transformation is a positive one, we must fix digital authentication.

THE UNCANNY VALLEY

When OpenAI launched ChatGPT, the age of AI began in earnest. People initially used ChatGPT for creative purposes, asking the bot to invent lyrics or write poems in iambic pentameter. It didn’t take long, however, for people to realize that AI was more than just a party trick. Since then, people have discovered new uses for it at home, work, and school. And today, you’re more likely than ever before to consume content created by an algorithm and not even know it.

Feeling creeped out yet? Welcome to the uncanny valley.

Although AI-powered chatbots are smarter than we could have imagined, they still have one major shortcoming: they can read a bit, well, robotic. When I asked ChatGPT to write an email about the growing popularity of the four-day workweek, it produced an eloquent missive about the importance of work-life balance in under a minute. Although I found the essay interesting, its tone fell flat. It wasn’t only that the tone didn’t sound like me, it was that it didn’t really sound like anyone. Reading it plunged me into the uncanny valley; It was impersonal and broad, lacking the idiosyncrasies that make good writing feel human. 

The American poet Wallace Stevens writes that “Style is not something applied. It is something that permeates. It is of the nature of that in which it is found, whether the poem, the manner of a god, the bearing of a man. It is not a dress.” To Stevens’ point, every formal element of writing, from syntax to word choice, reflects who a person is and where they come from. All of these tiny idiosyncrasies create a texture to writing that the glossy content written by AI lacks right now. In my experience, this lack of texture or, to use Stevens’ phrase, “personal style,” is AI’s only giveaway. That’s bound to change—and fast.

WHERE WILL AI GO FROM HERE?

It’s only a matter of time before AI shifts its gaze from the internet to the user. I believe that AI programs will soon analyze a user’s email and online correspondence to imitate their writing style, opinions, and worldview. Eventually, everybody could have their own unique AI that sounds just like them and shares their beliefs. The personalized AI could act as a digital avatar, saving “their human” time and energy by taking on tasks like responding to Slack messages, applying for home loans, and even preparing a legal defense for an upcoming trial.

This could have profound—and positive—effects on current society. Consider for a moment the fact that “54% of U.S. adults 16-74 years old—about 130 million people—lack proficiency in literacy, reading below the equivalent of a sixth-grade level,” according to the U.S. Department of Education. With the help of AI, millions of people lacking advanced reading and writing skills can gain more lucrative employment in the digital economy in unprecedented ways. Of course, there is also the dark side. In the next five years, I predict that AI will be the number one fraud threat in the world.

THE DARK SIDE OF AI

Imagine that it is 10 years in the future and your individualized AI was just hacked. The SMS OTP sent to your phone was intercepted by a fraudster via a SIM swap. Soon, your friends are receiving messages that sound just like you. This is not the plot of a sci-fi novel. This is what we are heading toward unless we can finally solve the internet’s central flaw: digital identity.

Today, fraudsters are winning the war on fraud. Every year, they bilk billions of dollars using established playbooks, imposing a major cost to governments, businesses, and individuals. As the world continues to become digital-first, the power of fraudsters will only grow. Unless changes to the way we authenticate and verify identity take place soon, the AI revolution could give fraudsters unprecedented power over their victims.

FUTURE-PROOFING DIGITAL AUTHENTICATION METHODS

Change can be difficult and phasing out familiar but flawed technologies (ahem, passwords and OTPs) is never easy, but here are three steps business leaders can take today to future-proof their digital authentication methods:

  1. By creating the digital equivalent of TSA’s pre-check line, you can cut down on wait times for legitimate customers while still maintaining stringent security measures for all.
  2. Use real-time phone signals (how long a phone has been in use, for example) instead of passwords to verify users. 
  3. Don’t make life any harder for your customers by requiring endless form-filling. Instead, leverage verified auto-fill data to accelerate onboarding while also reducing fraud.

As AI advances, we must recognize the challenges posed by the uncanny valley and the risks that could fuel the next generation of digital identity fraud. And while personalized AI can have positive effects, it will also come at a cost unless interested stakeholders create solutions that improve the state of digital identity authentication. I believe moving away from legacy knowledge-based forms of authentication is a crucial step toward preventing fraudsters from gaining unprecedented power.


Rodger Desai is the CEO & Co-Founder of Prove, the modern way of proving identity with just a phone.