{
  "id": 3154686,
  "title": "What would happen to human interaction if AI could read our emotions?",
  "url": "https://urgent.news/2026/08/25/what-would-happen-to-human-interaction-if-ai-could-read-our-emotions",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-25T01:00:02.000Z",
  "source": {
    "name": "Phys.org",
    "slug": "phys-org",
    "url": "https://phys.org/news/2026-08-human-interaction-ai-emotions.html"
  },
  "original_language": "en",
  "account": "Emerging AI systems demonstrate the ability to identify human emotions by examining facial expressions, vocal tones, body language, and physiological cues. While the scientific credibility of these technologies is still under debate, they are currently being tested in medical and security fields. Rather than concentrating on current limitations, PhD candidate Prégent poses an alternative question: what if these systems function flawlessly? Companies highlight the benefits of such technologies, but philosophers and scholars must highlight the potential risks if they prove reliable. To investigate the implications of fully functional emotion recognition technologies (ERTs), Prégent created several future scenarios. In one scenario, a surgeon approaches a hospital for surgery. Prior to entering the operating room, an ERT system identifies her as emotionally unstable. She is escorted by her employer and barred from operating. The surgeon asserts her professional competence, citing a recent divorce and concern for her child as the reasons for her emotional state. However, the technology has interpreted her emotional condition as a reason to doubt her ability to perform the procedure. Prégent highlights a broader societal concern. People engage in various social roles - surgeon, parent, partner, friend, and neighbor. They typically decide for themselves which aspects of these roles they bring to specific situations. ERT could blur these boundaries, as people may find their emotional lives exposed and compelled to explain or justify them, rather than deciding themselves what to reveal. Similar tensions may arise in public settings. Consider a government official delivering a public apology for a fatal incident. If ERT suggests the apology lacks genuine remorse or sadness, the public might focus on the discrepancy between the expressed emotions and the technology's interpretation. Public apologies are crucial, but they do not always reflect authentic emotions, such as in the case of police officers, doctors, or politicians dealing with traumatic events. According to Prégent, ERT could fundamentally alter how we interpret others. Instead of relying on people's actions and words, we might increasingly trust technology's emotional assessments. The European AI Act currently restricts some uses of emotion recognition technology due to its unreliability. Prégent's research explores this scenario if the technology becomes significantly more accurate. While ERT could aid in early detection of neurological disorders or assist individuals in recognizing emotions, these potential benefits should not overshadow the broader ethical concerns. Protecting emotional privacy extends beyond safeguarding information; it involves preserving the social environment where individuals can define their identity, present themselves, and engage with others. Prégent cautions that society must contemplate the consequences of ERT before companies decide on its implementation.",
  "summary": "AI systems are increasingly being developed to recognize emotions by analyzing facial expressions, voices, body language and other physiological signals. Although the scientific validity of many of these systems remains contested, they are already being tested in health care and security settings.",
  "key_points": [
    "ERT might blur personal boundaries, forcing individuals to justify their emotional lives.",
    "Society must consider ethical implications of ERT before widespread implementation."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}