Signal
Regulators are writing rules for machines that read feelings before the technology goes mainstream. The EU has banned AI that infers emotions at work and in education, the enforcement machinery arrives through 2026, and privacy lawyers are already mapping the boundary disputes: inference versus recognition, emotions versus intentions, and the medical and safety exceptions. The rules are being written while the field is still young.
So what?
Emotion-reading AI is being banned at work before it matures, which sends it to the one place the regulation does not reach: your home.
Evidence
UC Today, 20 April 2026 (news): Article 5 of the EU AI Act bans systems inferring emotions in workplaces and education, with fines up to EUR 35 million or 7 percent of global turnover. The ban has applied since 2 February 2025, and France's CNIL and Ireland's Workplace Relations Commission are among the named enforcers.
Future of Privacy Forum, 24 March 2026 (policy): FPF's analysis unpacks the prohibition's narrow scope, emotion inference from biometric data only, and flags open questions, including systems that detect intentions rather than emotions falling outside the ban.
Worth asking
Where is the line, for you, between a machine noticing how you feel and a machine acting on it?
