HumanTail Career Index
This edition, exposure rose fastest wherever a job is measured in units of output — tickets, tests, tickets closed, words, slides. It fell wherever someone has to answer for a decision. The gap between producing and deciding widened again.
Jobs gaining exposure this edition
Movement in our task-exposure estimate since the previous edition.
Test generation and self-healing regression suites moved from pilots into standard delivery pipelines.
Search demand shifted toward AI answer surfaces, cutting volume commissioning further.
Tier-1 resolution rates crossed the threshold where teams stop backfilling attrition.
Screening and scheduling now ship inside standard ATS products.
Pod sizes for ticket-scoped delivery shrank while billing targets held.
Variant and resize work is fully generated inside design tools.
Long-document synthesis quality removed most secondary-research hours.
Jobs getting safer
Exposure fell because the human-held part of the work grew relative to the automatable part.
Documentation relief gave back hours without touching the hands-on work.
Relationship-held revenue kept resisting automation across FMCG and BFSI channels.
Documentation automated, but physical diagnosis and sign-off stayed with people.
Compliance production automated faster than advisory demand fell — advisory hours per client rose.
Prep automation improved the job; classroom accountability is unchanged.
Skills rising fastest
Ranked by how much they change an individual's exposure, not by course popularity.
Every team now needs someone who rebuilds the team's process around AI instead of buying another tool.
Accountability for what ships is the new junior job description.
As production costs fall, the paid part of the work becomes the conversation.
Regulated judgment cannot be delegated to a model.
People who own a number survive reorganisations that remove people who own a task.
It became table stakes, not a differentiator. Do not build a career on it.
AI capability improvements behind these moves
We only move a score when a capability becomes reliable enough for ordinary teams to depend on it.
Freshers and industries
Entry-level remains the sharpest edge. Roles whose training period existed to produce cheap output are the ones shrinking. Roles whose training period exists to build judgment are still hiring.
IT services, BPO/KPO, support and content production account for most of the upward movement this edition. Healthcare, field sales, site engineering and education account for most of the downward movement.
Cite this analysis
These figures are free to quote with attribution to HumanTail. Suggested citation:
HumanTail, "HumanTail Career Index — September 2026", September 2026 edition. https://humantailai.com/career-index
Press and research queries: press@humantailai.com
How the numbers are built
- We score tasks, not job titles. Each role is decomposed into the duties people actually perform in India, and each duty carries an exposure weight for how well current AI systems do it end to end.
- Role exposure is the weighted duty score, adjusted for the Indian delivery context: how the work is priced (per task, per ticket, per document), how much of it is screen-based, and how much requires physical presence or accountable sign-off.
- City, industry and salary indices are the weighted exposure of the local job mix — not measured job losses. Salary is used as a proxy for scope: the further a role sits from owning a decision, the higher its exposure.
- Every score carries a confidence label (HIGH, MEDIUM, LOW) based on how much verified duty-level data supports it. Low-confidence numbers are labelled, not hidden.
- Figures are estimates published for public scrutiny. They are recalibrated every month against reported hiring, headcount and capability data.
Published data we calibrate against
Third-party figures — sourced and linked. Our scores are estimates built on top of them.
Net headcount change across India's five largest IT services firms in FY26, against +12,718 in FY25.
CALIBRATES: IT services role exposure and the 'RISING' trend flag on delivery roles
Fresher hiring in Indian IT fell from a peak of about 600,000 in FY22 to roughly 120,000 in FY25 — an 80% decline.
CALIBRATES: Entry-level and student exposure; the ₹3–6 LPA bracket index
Entry-level IT roles in India have already shrunk by 20–25% as automation absorbs testing, basic coding and documentation.
CALIBRATES: Manual QA (86) and IT services engineer (74) exposure estimates
Tech openings for professionals with up to two years of experience fell to 10,000 from 13,000 year on year.
CALIBRATES: Fresher reality notes on every role card
India's technology workforce grew just 2.3% in FY26, to 59.5 lakh — its weakest expansion in years.
CALIBRATES: Sector-wide pressure forecast for IT services
Bengaluru, Delhi-NCR and Mumbai hold 25.4%, 24.8% and 19.2% of India's AI-related job openings — nearly 70% between them.
CBRE Research (64,500+ Naukri listings)
CALIBRATES: City exposure indices and AI-demand concentration column
64% of new Global Capability Centre roles in India now require AI, data science or automation skills; tier-2 hiring is growing 23% year on year.
foundit Insights Tracker, via FlowVerify
CALIBRATES: Tier-2 city index and the skill-ROI ranking
Workers with AI skills command a 62% average wage premium globally, and AI-skill roles are growing about eight times faster than the wider market.
CALIBRATES: Skill ROI ordering and salary-at-risk framing
India recorded a 59.5% year-on-year jump in AI engineering hiring — the fastest of any major economy.
CALIBRATES: Human Moat uplift for AI-adjacent roles
Peer-reviewed Indian research on AI's labour-market effect, used as a sanity check on our sector-level ordering.
ICRIER — 'AI and Jobs: This Time Is No Different'
CALIBRATES: Industry index ordering across all 23 sectors
This index is an average. You are not.
Two people with the same job title can sit twenty points apart depending on what they actually do all day. Get your own number instead of reading ours.