👁️ The Disturbing Rise of Unethical AI
AI tools explicitly designed to cheat, invade privacy, trick people, and violate rights are on the rise. We seem to be entering a new and unsettling phase in AI | Edition #203
👋 Hi, Luiza Jarovsky here. Welcome to our 203rd edition, now reaching 61,400+ subscribers in 168 countries. To upskill and advance your career:
AI Governance Training: Join my 4-week, 15-hour program
Learning Center: Receive free AI governance resources
Job Board: Find open roles in AI governance and privacy
Subscriber Forum: Participate in our daily discussions on AI
☀️ Summer Training
If you want to upskill and explore AI's legal and ethical challenges in an in-depth and interactive way, I invite you to join cohorts 21 or 22 of my live online AI Governance Training, a 4-week program starting in mid-July and early September.
Over 1,200 professionals have participated, and many have described the experience as transformative (testimonials here). Today, we're launching an early bird discount for the summer cohorts (valid until Friday). I hope to see you there!
👁️ The Disturbing Rise of Unethical AI
In recent weeks, I've encountered new consumer-facing AI tools that are explicitly designed for cheating, invading privacy, tricking people, and ignoring rights. No fluffy marketing: just deception and disregard for personal and ethical boundaries.
In today's edition, I discuss examples of unethical AI, the unsettling new phase we seem to be entering in AI, and its legal implications.
1. Cluely
Cluely is marketed as “invisible AI to cheat.” Its CEO posted a promo video and invited people to cheat on everything.
The startup was initially presented as an AI tool to cheat on job interviews. The founder was suspended from Columbia University but still raised $5.3 million in seed funding.
In the company's manifesto, it states: “So, start cheating. Because when everyone does, no one is.”
2. Gigi
Gigi is marketed as “the AI who knows everyone.” On its homepage, it adds:
“I browsed the internet while you were busy living.
Your tweets. LinkedIn. Substack. Insta.
I turned it all into a profile: the one that actually gets you.
You didn’t sign up.
I just knew you were worth mapping.”
At the bottom, it lists “the hot list,” and something tells me they took inspiration from Zuckerberg's Facemash (when Zuckerberg was a college student, he hacked Harvard's servers to download students’ pictures and upload them on Facemash so others could vote for the most attractive).
Today, Zuckerberg calls Facemash a “prank website” and distances himself from it. Gigi, by contrast, unapologetically embraces the creepy branding.
3. Mechanize
For now, this single screen is all Mechanize has, but it's backed by well-known figures in the tech world. The company was launched last month, and its founder announced:
“Mechanize will build virtual work environments, benchmarks, and training data to enable the full automation of all work."
They want to explicitly automate all human labor and absorb all funds that currently go to salaries.
-
The three AI startups above were recently launched and are proudly, openly unethical.
For them, cheating, invading people's privacy, and automating all human work are seen as fair and legitimate goals, with no additional considerations, disclaimers, or justifications needed.
They chose this bold and unapologetically unethical messaging on purpose, as they knew there would be demand for such AI tools and services.
The problem goes beyond unethical AI startups: