# TCS CEO Krithivasan backs AI, says revenue cannibalisation is fine as younger staff outpace seniors

> TCS is leaning into AI, even if it trims services revenue. The bet: faster delivery, outcome-based deals, and upskilled teams beat billable hours.

- Type: AI news
- Page: https://completeaitraining.com/news/tcs-ceo-krithivasan-backs-ai-says-revenue-cannibalisation/
- Published: 2026-02-26
- Updated: 2026-02-26
- Author: Jeroen Erne
- Jobs: Executives and strategy
- Reading time: 3 min

### TCS is fine with AI eating its own revenue. Here's the executive playbook behind that stance

Tata Consultancy Services signaled two things leaders should pay attention to: they're not afraid of AI, and they're comfortable with revenue "cannibalisation." The company also sees a clear skill gap - younger employees are faster at building AI solutions than senior staff.

That's not a PR line. It's a strategic posture. If the largest IT services firm is willing to disrupt its own billable hours with AI, the market will reward speed and outcomes, not bodies and time.

#### Why this matters now

- Clients expect AI-driven productivity. If you don't offer it, someone else will.
- Cannibalisation is a pricing and delivery shift, not a revenue death sentence.
- The seniority skill gap is solvable with structure, incentives, and hands-on practice.

#### Treat cannibalisation as a feature, not a bug

Revenue will move from time-and-materials to outcome, platform and subscription-like models. That is healthy if your unit economics improve.

- Set a cannibalisation threshold: accept X% services revenue reduction in exchange for Y% margin expansion and Z% win-rate lift.
- Bundle AI accelerators with delivery. Reuse internal components to compress cycle time and defend price.
- Shift proposals to outcomes and SLAs (e.g., cost-to-serve cuts, cycle-time targets), not hours.

#### Close the seniority skill gap fast

TCS notes seniors lag while younger staff move faster with AI. That's common - senior talent is optimized for governance and client context, not new tooling.

- Set a 6-8 week, role-based upskilling plan with weekly build targets (working demos, not slideware).
- Pair senior architects with "AI-native" developers in two-pizza squads. Rotate lead roles every sprint.
- Make AI capability part of performance and compensation. Reward shipped AI features and reusable assets.
- Institutionalize reverse mentoring. Younger "proficient" staff teach workflows; seniors stress non-functional requirements and client fit.

#### Productize your delivery

Move from bespoke projects to repeatable systems.

- Create an internal AI platform team to manage models, prompt libraries, evaluation harnesses, and guardrails.
- Publish a catalog of proven patterns: retrieval-augmented generation, code copilots, test automation, document extraction, and customer support assistants.
- Standardize toolchains, data connectors, and evaluation metrics so every new deal starts at 60% complete.

#### Pricing and contracts that work with AI

- Offer a "productivity rider" that guarantees measurable efficiency gains in exchange for committed scope.
- Use tiered pricing: discovery (fixed), build (milestone-based), operate (subscription/consumption with uptime and quality SLAs).
- Separate one-time enablement from ongoing optimization to avoid hiding platform value in services lines.

#### Governance without slowing delivery

- Define model risk classes (public, fine-tuned, custom) and approval paths for each.
- Automate evaluations: accuracy, toxicity, PII leakage, and cost-per-output. Ship with red-team signoff.
- Lock in data controls: masking, retrieval scopes, and clear IP ownership in contracts.

#### The metrics that matter

- Delivery: cycle time, defects escaped, cost-to-serve, reuse rate of AI assets.
- Commercial: attach rate of AI components, margin mix, renewal/expansion tied to outcomes.
- People: percentage of senior staff shipping AI features, time-to-first-deployment, pair-programming adoption.

#### 90-day executive action plan

- Week 1-2: Pick three high-volume workflows to automate 30-50% (e.g., test generation, requirements summarization, L2 support).
- Week 3-6: Stand up two cross-functional squads. Deliver internal pilots with clear before/after metrics.
- Week 7-10: Productize the best patterns into reusable components. Publish the internal catalog.
- Week 11-13: Update pricing templates and SOW language to reflect outcomes and AI accelerators. Train sales.

#### Bottom line

TCS is telling the market: we'll trade some services revenue for speed, margin, and relevance. That's the rational move when AI compresses effort and clients care about outcomes.

Your edge won't come from "adopting AI" as a slogan. It will come from incentives, reuse, and contracts that make doing the right thing the easy thing.

[AI for Executives & Strategy](https://completeaitraining.com/tag/executives-and-strategy/)

---

This is the Markdown form of https://completeaitraining.com/news/tcs-ceo-krithivasan-backs-ai-says-revenue-cannibalisation/. The whole site is mapped at https://completeaitraining.com/llms.txt.
