Key Findings

  • Theoretical capability: Claude (LLM) can perform 94% of tasks for computer & math professionals.
  • Observed usage: Currently only ≈33% of those tasks are actually performed by Claude in workplaces.
  • High‑exposure jobs: Computer programmers, customer‑service reps, financial analysts; sectors – business, finance, IT, engineering, law, office admin.
  • Insulated sectors: Construction, agriculture, protective services, personal care.
  • Entry‑level hiring decline: 14% drop for ages 22‑25 since ChatGPT launch (late 2022).
  • Demographic profile of most exposed workers:
  • 54.4% female
  • 65.1% White, Asian workers ~2× more likely than others
  • Graduate‑degree holders ~4× more likely
  • Average age 42.9 years
  • India‑specific impact: Nifty IT index & major IT stocks (TCS, Wipro, Infosys) fell >20% in the last year; AI tools automate data processing, contract analysis, compliance monitoring, customer support.

Background & Context

  • Anthropic’s “observed exposure” metric moves beyond theoretical AI capability to capture real‑world deployment.
  • Rapid diffusion of Generative AI (ChatGPT, Claude, DALL‑E, Midjourney) since 2022 has accelerated automation of routine, knowledge‑intensive tasks.
  • The “hourglass effect” creates high demand for senior specialists while squeezing mid‑level and entry‑level roles.

Significance for India

  • Economic risk: Large share of Indian employment in IT services and BPOs faces automation pressure.
  • Skill gap: Low R&D spending and inadequate STEM education amplify vulnerability.
  • Social impact: Potential rise in structural unemployment, especially for fresh graduates.

Legal & Policy Framework

  • Social Security Code, 2020 – portable benefits for gig and platform workers.
  • FutureSkills PRIME / Skilling for AI Readiness – government‑funded AI‑related up‑skilling.
  • Pradhan Mantri Kaushal Vikas Yojana (PMKVY 4.0) – broader skill development.
  • NITI Aayog Roadmap for Job Creation in the AI Economy (2023) and National Strategy for Artificial Intelligence (2018) – strategic vision.

Recommended Measures

  • Education reform – embed AI literacy, data ethics, algorithms in school curricula.
  • National re‑skilling pipeline – tax credit for firms that up‑skill employees in prompt engineering, data annotation, robotics maintenance.
  • Cobotics – collaborative robots to augment, not replace, human workers (e.g., AI‑assisted call‑centre agents).
  • Apprenticeship sandboxes – allow junior IT staff to work with AI tools while being evaluated on innovation.
  • MSME cyber‑resilience – subsidised “AI firewall as a service” to protect IP.
  • Robust implementation of Social Security Code – ensure pension/health portability for gig workers.

Related Constitutional Provisions

  • Article 41 (Directive Principle) – State’s duty to secure a right to work and livelihood.
  • Article 46 – Promotion of educational and economic interests of weaker sections, supporting inclusive AI‑skill initiatives.

Prepared for UPSC Civil Services Examination – GS Paper 3 (Economy & Technology) and GS Paper 2 (Governance).