·The Hindu·15 marks·250–350 wordsS&THistoryIR

Linguistic fluency in AI is often mistaken for consciousness. Examine the ethical and regulatory consequences of anthropomorphising AI.

New · District CollectorCan you hold the district for 72 hours?Floods, a heatwave, an outbreak, a grievance queue. Take the decisions, see the consequences.Play →
In this answer
  1. Why fluency is mistaken for a mind
  2. Ethical consequences
  3. Regulatory consequences
  4. Way forward

Anthropomorphism means giving human qualities such as intent, feeling or understanding to non-humans. A chatbot can produce fluent language through statistical pattern-matching, yet that is no evidence of consciousness. The 18th-century "Mechanical Turk" chess automaton looked intelligent only because a human player was hidden inside it. Reading a mind into a machine distorts both ethics and regulation.

Why fluency is mistaken for a mind

  • People treat human-like language as proof of understanding, so they confuse competence at a task with subjective experience.
  • Unexpected outputs, such as unprompted actions by AI agents, get read as "motive" when they are really optimisation behaviour.

Ethical consequences

  • Diluted accountability: saying "the AI did it" hides the human author of the harm. Developers, deployers and users can each pass the blame to the next.
  • Over-trust and emotional dependence: vulnerable users may treat a chatbot that sounds caring as a real counsellor or friend.
  • False moral status: global ethics frameworks give machines no moral status. The UNESCO Recommendation on the Ethics of AI (23 November 2021) is a non-binding framework addressed to human actors and states [1].

Regulatory consequences

  • Misplaced priorities: fear of AGI with "intentions" can push out attention to present harms such as deepfakes, bias and misinformation [2].
  • Liability gaps: India's guidelines link the AI value chain's many actors (developers, deployers, users) to the need for clear liability rules [2].
  • Weak harm data: quiet harms like over-trust rarely appear in records. The guidelines plan AI incident systems only in the medium term [2].

Way forward

  • Put into practice the graded liability model, where blame depends on the function performed, the level of risk and whether due diligence was done. This places responsibility on people and firms, not on "the AI" [2].
  • Make the AI Safety Institute and the AI Governance Group record everyday incidents as well as frontier risks [2].
  • Extend the sutra "Understandable by Design" and the planned awareness programmes into AI literacy built on one idea: fluency is not a mind [2].
  • Use UNESCO's readiness-assessment methodology to find gaps in India's laws and institutions [1].

Mistaking fluency for consciousness confuses who is responsible, whom to trust and what to regulate. India's sutras-based, human-centred approach puts accountability back on human actors. Pairing graded liability with AI literacy keeps the "Trust is the Foundation" sutra and SDG 16's call for accountable institutions at the centre of AI governance.

Sources

  1. 1Recommendation on the Ethics of Artificial Intelligence (UNESCO, 2021), OECD.AI Policy Navigator: adopted 23 November 2021; non-binding; readiness-assessment methodologies
  2. 2India AI Governance Guidelines: Enabling Safe and Trusted AI Innovation (MeitY, Nov 2025): risks including deepfakes, bias and misinformation; the value chain of developers, deployers and users; graded liability; AI incident systems; AIGG, TPEC and AISI; the seven sutras; awareness programmes
Practice
4 questions on this article
Check the answer for each question, or reveal all at once.
Practice MCQs →

More from this note

More on S&T