Emotional machines and trust: the privacy and manipulation problem

October 6, 2026 · 20 min read

Kate Darling warns about robots as marketers, Rana el Kaliouby addresses bias and trust in emotion AI, and Lex Fridman lays out his hopes for companion AI. Anyone weighing a social robot or companion app can see here both the promise and the specific ways the emotional bond could be exploited.

12 questions on this insight. Answer in your head, then open the card.
What do Fridman, Darling and el Kaliouby agree on about people and machines? Show answer

People bond with machines that show even a few social cues, and that cannot be trained away, so products must be designed around it.

What safeguards do all three support in principle? Show answer

User control of data, clear opt-in and trustworthy leadership. Fridman and Darling also prefer a product you buy over one funded by advertising.

What does Lex Fridman want from a companion machine? Show answer

One that remembers shared moments, belongs to you, and works for your long-term growth and happiness instead of your engagement.

What is Fridman's test for a free relationship with a machine? Show answer

You own your data, can see how it is used, and can delete all of it and walk away.

Why does Kate Darling call a social robot a potent marketing tool? Show answer

It combines personalized recommendations, social persuasion, automation, scalable data collection and a physical presence.

Why does Darling say consent is not enough? Show answer

Data aggregated from many people allows inferences about individuals who never shared anything, and it is often used against vulnerable groups.

What did the study of children and social robots find? Show answer

Most children preferred that robots market through casual conversation, because the robot knows them, without realizing it acts for the company.

What is Rana el Kaliouby's biggest worry about AI? Show answer

Bias built into systems that are deployed at scale, not a takeover.

What can emotion-recognition technology honestly claim, according to el Kaliouby? Show answer

It can measure what shows on the face. A smile or a furrowed brow has many causes, so it cannot claim to know true feelings.

What values did Affectiva set at its founding? Show answer

Respect privacy, require clear opt-in, and avoid security and surveillance work.

What does the Jibo story show? Show answer

The bond is real enough to hurt. El Kaliouby's son grieved when the robot shut down after its company ran out of money.

Where do Fridman and Darling disagree? Show answer

Fridman thinks ownership, deletion and opt-in solve the problem. Darling says they protect one informed adult and leave inference, shaped wants and children exposed.

This is the short form. The full insight has every expert's reasoning, their own words and the sources.