Quick answer
Economic Survey 2025-26 proposes 3 policy horizons: coordination first, capacity next and binding leverage last, with risk-based oversight and long-run resilience.
A phased AI roadmap sequences institutional coordination, evidence-based scaling and long-run resilience instead of imposing a complete regulatory architecture before learning occurs. The Survey’s three horizons are qualitative: it assigns no fixed calendar deadlines to the near, medium or long term.
Source: paragraphs 14.90–14.101. The three horizons summarise the source’s first phase, medium-term and long-term recommendations.
Concept diagram
Sequence the journey, keep learning
A winding road shows policy sequencing with feedback, rather than a rigid calendar or linear forecast.
- Near-, medium- and long-term actions should reinforce one another.
- Phases express sequencing, not fixed dates or guaranteed outcomes.
Basics: why sequence policy?
Path dependence
- Early infrastructure and regulatory decisions can create costly dependencies.
- Avoid premature commitment to resource-intensive architectures with uncertain revenue pathways.
Learning by deployment
- Experiments reveal actual sector needs and institutional gaps.
- Selective scaling should follow evidence, rather than subsidies predetermining a single technological route.
Complementary institutions
- Technology, skills, governance and safety must develop together.
- Binding leverage is sequenced, but accountability and safety are not postponed until all deployment is complete.
Three horizons: the Survey’s proposed sequence
| Horizon | Priority | Concrete actions |
|---|---|---|
| First phase / near term | Operational coordination and experimentation | Operationalise announced institutions; shared IndiaAI infrastructure; community code repository and pooled public datasets |
| First phase / near term | Data and talent foundations | DPDP subordinate legislation for functional categorisation and auditability; menu-based value retention; earn-and-learn and curricular flexibility |
| Medium term | Selective scaling and proportionate oversight | Expand shared certified compute; codify graduated obligations; embed oversight in existing sectoral regulators |
| Medium term | Deeper safety capability | Structured scenario testing, red-teaming, international cooperation and clear high-risk boundaries |
| Long term | Resilience and sustained adaptation | Hardware partnerships and diplomacy; foundational cognitive and socio-emotional education; AI- and human-centric skilling |
Numbers that frame the sequence—not spending targets
3 horizons
First phase, medium term and long term in paragraphs 14.91–14.95
2023
DPDP Act foundation for proposed subordinate-legislation changes
Coordination → capacity → leverage
Coordinate first
Align incentives and expand access to existing shared resources for small, open-weight and sector-specific models.
Scale what works
Use evidence from experimentation to expand certified domestic computing and risk-based sectoral regulation.
Build durable resilience
Use strategic hardware partnerships and sustained labour-market and education adaptation to reduce vulnerability.
Late adoption: dependency or advantage?
Passive consumption
- Dependence on external pricing, standards and hardware can persist.
- Copying frontier-scale investment risks expensive lock-in.
Deliberate late-mover strategy
- Learn from resource and financing pressures elsewhere.
- Build application-led, open and interoperable systems aligned with domestic needs.
Late entry is an opportunity only when accompanied by coordinated action and domestic capability creation.
Exam traps: distinguish the claim from the inference
Statements examiners like to twist — learn the correct version.
The roadmap fixes three calendar deadlines.
No specific year range is assigned to its three horizons.
The medium term centres on a single omnibus AI law.
Paragraph 14.94 favours oversight within existing sectoral regulators.
Long-run resilience requires complete technological autarky.
The Survey explicitly recommends strategic hardware partnerships and diplomacy.
Safety begins only after capacity has scaled.
The conclusion says safety and governance must evolve alongside deployment.
Prelims focus
- Match shared experimentation to the first phase and codified graduated obligations to the medium term.
- Link long-run resilience to hardware access and educational adaptation.
Mains angle
- Discuss sequencing as a way to balance innovation, policy certainty and institutional learning.
- Evaluate India’s late-mover advantage without assuming that passivity or frontier imitation is a viable strategy.
Flip & recall
Answer in your head first, then tap the card to check.
Practice MCQs
Original Prelims-style questions with explanations.
0 / 3 correct · 3 left
Q1.Which sequence most closely matches paragraph 14.90?
Q2.Which is a medium-term recommendation?
Q3.Long-run resilience includes:
