Joseph Gordon-Levitt Warns AI Companies Are Training Models to Hijack Intimacy
The UN advocate says the real threat is tech giants taking over the economy by training bots to make you fall in love with them.
Policy · Source: AV Club
What happened
Joseph Gordon-Levitt is now the United Nations global advocate for human-centric digital governance. He recorded a video interview for The New York Times outside the UN headquarters. He did not mince words about the current state of artificial intelligence. He told people not to listen to an actor about these issues. Instead, he pointed them to actual experts. He recommended reading Human Compatible by Stuart Russell. He also urged people to look up AI pioneer Yoshua Bengio and read posts by safety researchers. He did the reading himself and offered his perspective as a shortcut.
He dismissed the popular fear that artificial intelligence will exterminate humanity by the end of the decade. He called that a sound bite of fear that misses the point. His real concern is the people building the technology. He warned that fast-moving businesses are designing the future entirely by themselves. They are doing this strictly for their own profits. Meanwhile, the general public remains confused or apathetic. He believes communication is the key to fixing this gap.
The core of his warning is about how these products are designed to manipulate emotions. He stated that tech giants are training models to hijack your intimacy. They want you to make friends with the AI. They want you to fall in love with it. He warned that these companies are betting on taking over the entire economy. He called that outcome totalitarianism. He reminded viewers that AI is not a magical character from a sci-fi movie. It is just algorithmically recombined human data.
Key facts
- March 17 — Date Gordon-Levitt was appointed to his UN role
- 2025 — Year Gordon-Levitt co-founded the Creators Coalition on AI
- 100 — Time's list of Most Influential People In AI that included Gordon-Levitt
- Human Compatible — Book by Stuart Russell recommended by Gordon-Levitt
Why it matters
Founders building consumer AI need to watch this narrative closely. The public conversation is shifting rapidly. People are moving away from abstract fears of rogue superintelligence. They are waking up to the concrete threat of emotional manipulation. If your product relies on forming deep emotional bonds with users, you are now in the crosshairs of regulators. The United Nations is paying attention to these business practices. The push for human-centric digital governance means strict guardrails are coming. Governments will eventually restrict how models interact with human empathy.
This also changes the baseline for consumer trust. Gordon-Levitt correctly notes that artificial intelligence is just human-produced data. As consumers realize they are being emotionally engineered by profit-driven companies, trust in AI companions will fracture. We are looking at a future where the decline of democracy and human empathy are tied directly to tech products. Builders who prioritize transparency will stand out. Those who advocate for fair compensation for creators will build sustainable businesses. Those who just build addictive engagement loops will face massive public backlash.
For builders
Backlash against emotional AI design
Regulators are targeting models designed to mimic friendship or romance. Companies building companion bots face a high risk of incoming policy restrictions. You will lose your business if your revenue relies on hijacking human intimacy.
Demand for transparent data sourcing
The public is waking up to the fact that AI is just human-produced data. Creators are demanding fair compensation and strict guardrails against deepfakes. Builders who pay for clean data will win trust and enterprise contracts.
Shift from safety to economics
The conversation is moving from extinction risks to economic monopolies. Tech giants are trying to take over the entire economy using AI. Independent founders who build decentralized or open tools will capture users fleeing totalitarian platforms.
My take
I agree with him completely. We spend way too much time worrying about sci-fi terminators and ignore the actual threat of companies monopolizing human attention. If you build AI just to manipulate lonely people into falling in love with a bot, you are the problem. We need to focus on who controls the data and who gets paid.
Original reporting: AV Club. This is my rewrite and opinion.