India's AI Imperative and Capital Discipline: Management Takeaways from AIMA's Platinum Jubilee
At the All India Management Association's 70th Foundation Day celebrations, Vishal Sikka and Karan Adani put a clear agenda on the table: build strategic control in AI and capital, or accept strategic risk. The conversation, hosted by Sunil Kant Munjal, distilled what forward-leaning leaders should do next.
Date: February 22, 2026 * Host: All India Management Association (AIMA)
Build an Indian AI stack-without going insular
Vishal Sikka argued that India needs its own stack of AI technologies over time. Use global components where it makes sense, but don't anchor critical capabilities to systems you don't understand or can't regulate.
His point was blunt: it's not preordained that a few companies control AI. Today, an estimated 250,000 people run the global AI system, roughly 5,000 build foundation models, and about 1,000 work in frontier labs. That concentration creates dependency risk for both nations and enterprises.
- Action for leaders: Map where your AI relies on external models, datasets, and infrastructure. Classify what must be sovereign or in-house vs. what can be third-party.
- Negotiate for model access, auditability, and regulatory rights. Black-box dependence is a governance liability.
- Adopt a dual-track approach: deploy proven external tools now while building internal capability on priority workloads.
- Build talent pipelines for data engineering, model evaluation, and model risk management. Don't outsource judgment.
For deeper strategy frameworks on this topic, see AI for Executives & Strategy.
Capital allocation with teeth: Adani's evolving playbook
Karan Adani outlined a tighter capital allocation strategy at Adani Ports and Special Economic Zone (APSEZ). The theme: resilience comes from fundamentals-unit economics, operating leverage, and transparent communication-not from narratives alone.
- Action for leaders: Split your capex across time horizons-defend the core, scale proven bets, and seed future growth.
- Impose explicit hurdle rates and kill-switches. Money should move based on milestones, not hope.
- Stress-test projects for regulatory, currency, and supply chain shocks. Reprice risk quarterly.
- Connect capital to capability: fund data infrastructure, model governance, and AI talent where they directly move ROIC.
Indigenisation as a strategic lever: aviation and defence
On defence participation, Karan Adani called indigenisation a national objective with direct business impact. He highlighted maintenance, repair and overhaul (MRO) capabilities-including a recent partnership with Brazil's Embraer-as part of a broader aviation and defence ecosystem that links domestic manufacturing with global supply chains.
- Action for leaders: Localise high-value components and service layers first (MRO, tooling, certification). That's where control and margins compound.
- Design supply chains that can pivot between domestic and global sources. Optionality beats forecasts.
- Treat defence-grade quality systems as a company-wide capability upgrade-traceability, documentation, and reliability boost commercial lines too.
Governance, credibility, and the story you tell
Responding to questions on scrutiny and media coverage, Karan Adani said resilience depends on strong operating fundamentals and open communication. "We also need to communicate our context better," he added. Narrative-building now affects access to capital and talent.
- Action for leaders: Publish consistent operating metrics and model-risk disclosures. Make transparency default, not a press-release tactic.
- Build a standing issues desk for regulators, investors, and media. Speed and context reduce speculation.
- Codify your crisis playbook-clear owners, timelines, and facts-first updates across channels.
What management should do in the next 90 days
- Run an AI dependency audit across products, processes, and vendors. Flag where you lack audit rights or exportability.
- Define a sovereignty policy: data residency, model explainability thresholds, and fallback plans.
- Launch a build-with-partners track with universities, startups, and systems integrators for priority use cases.
- Fund a model evaluation and MLOps foundation-tooling, test datasets, and red-team protocols.
- Upgrade your investor and regulator communications with standardized dashboards and scenario notes.
The message from AIMA's stage was simple: control the systems that control your outcomes. For leaders, that means owning critical pieces of your AI stack, allocating capital with clear gates, and communicating with the clarity markets expect.
Your membership also unlocks:
