AI Legal Drafting for Indian Startups: From Term Sheet to Employment Letter

By LegalInk Editorial ·

For Indian startups, AI drafting is most useful across four document classes: term sheets and SHAs, ESOP grant letters, employment agreements, and high-volume vendor and NDA paperwork. These documents are structurally repetitive but legally consequential — the right fit for template-driven AI drafting paired with focused counsel review. The workflow below shows where each tool fits and which clauses still need a human read.

Why the Startup Document Stack Suits AI Drafting

A funded Indian startup generates concentrated bursts of legal paperwork — at each funding round, each senior hire, each regulatory milestone. The documents follow settled structures in Indian practice but carry real downstream risk if mis-drafted.

A typical Series A-ready document stack includes:

  • Term sheets and investor offer letters
  • Shareholder Agreements (SHA) and Share Subscription Agreements (SSA)
  • ESOP scheme rules and grant letters
  • Employment agreements, offer letters, and NDAs
  • Vendor and SaaS agreements

AI drafting tools can hold the structure, apply the governing law clause correctly, and surface missing provisions. In-house counsel then spends time on negotiation points, not on rebuilding boilerplate.

Stage 1: Term Sheets and Investment Documents

What the term sheet must anchor

A term sheet is non-binding on economics but binding on exclusivity and confidentiality in most Indian deals. When reviewing an AI-drafted term sheet, verify that the exclusivity window, no-shop clause, and conditions precedent list are explicitly flagged as binding or non-binding, as the case requires.

The Companies Act, 2013 governs the downstream share issuance. SHA provisions dealing with anti-dilution, drag-along, or tag-along rights must align with the Articles of Association, since the Articles govern what is enforceable against third parties. An AI drafting tool configured for Indian practice should prompt this AoA-consistency check automatically.

SHA and SSA: where AI drafting reduces risk

Shareholder agreements in Indian startup deals often import VC-standard clauses — liquidation preferences, information rights, board composition rights — without adapting them to Indian corporate law constraints. AI drafting adds value by flagging clauses that may conflict with mandatory Companies Act provisions or SEBI (ICDR) Regulations where a public-facing exit is contemplated.

Foreign investment adds another layer. FEMA 1999 and the Foreign Exchange Management (Non-Debt Instruments) Rules, 2019 govern pricing and reporting for foreign equity infusions. Any SHA or SSA involving a non-resident investor should treat FEMA compliance as a mandatory checklist item. Run draft investment documents through the LegalInk compliance framework for a FEMA and RBI reporting check before they go to investor's counsel.

Stage 2: ESOP Documents

ESOP rules under Companies Act, 2013

ESOPs for private companies are governed by the Companies Act, 2013 read with the Companies (Share Capital and Debentures) Rules, 2014. The scheme requires a special resolution of shareholders. Grant letters issued under the scheme must reference the scheme document and specify grant date, exercise price, vesting schedule, and lapse conditions.

AI drafting works well for grant letters because the structure is highly standardised and the variables — grant date, number of options, vesting cliff, exercise price — are data fields, not judgment calls. Maintain a master grant letter template, push updated variables at each grant cycle, and generate a batch for board approval in a single session.

What to review manually

The lapse-on-termination clause requires human judgment. Whether unvested options lapse immediately on termination for cause, and whether vested options survive for a defined exercise window, is a negotiated term that varies by seniority. AI-drafted text will produce a defensible default — but apply the company's HR policy and any specific hiring commitments before finalising.

Stage 3: Employment Agreements and Offer Letters

Governing framework

Employment in Indian startups is governed by a patchwork: the Industrial Employment (Standing Orders) Act, 1946 where applicable, the Shops and Establishments Act of the relevant state, the Code on Wages, 2019, and the Code on Social Security, 2020. The four Labour Codes — not yet uniformly notified across states — are the direction of travel, and any agreement drafted now should anticipate the definitions and compliance obligations under the Codes.

AI drafting is well-suited to generating state-specific employment agreements that reflect the correct Shops Act registration state, the applicable gratuity and PF contribution structure, and IP assignment clauses that hold up under Indian judicial reading.

Non-compete and confidentiality

Indian courts apply Section 27 of the Indian Contract Act, 1872 strictly: post-termination non-compete clauses are generally unenforceable as restraints of trade. A well-drafted employment agreement should focus on confidentiality obligations, non-solicitation of clients and employees (which courts treat more permissively), and a robust IP assignment clause capturing inventions made during employment.

AI-drafted employment agreements in the Indian context should default to this position — not the broad post-termination non-compete that appears in many imported templates. When reviewing AI output, check that the non-compete is either absent or limited to the employment period.

Offer letters at scale

For startups making 20–50 hires a quarter, the offer letter is a drafting-volume problem. The LegalInk assistant supports structured generation for repeating document types — offer letters, NDAs, vendor agreements — where the core terms are consistent and the variables are role-specific.

Building a Review-Light Workflow

The goal is to move counsel effort from drafting to review, and from review of boilerplate to review of negotiated terms. A functional workflow for a Series A-stage startup looks like this:

  1. Template library — Maintain approved master templates for each document class, signed off by external counsel annually or on major statutory change.
  2. Variable population — AI drafts populate deal-specific or hire-specific variables against the approved template.
  3. Exception flagging — The tool flags deviations from the approved template or missing mandatory provisions.
  4. Counsel review — In-house counsel reviews flagged exceptions and negotiated terms only.
  5. Compliance check — Run a FEMA, labour law, or sector-specific compliance check before execution.

This structure reduces per-document time substantially and creates an audit trail of what changed from the approved baseline — useful if a document is disputed later.

Frequently Asked Questions

Can AI-drafted SHA clauses be enforced under Indian law?

Yes, provided the clauses are consistent with the Companies Act, 2013 and the company's Articles of Association. Enforceability issues arise from inconsistency with mandatory statutory provisions or the AoA, not from how the draft was generated. Review for AoA alignment before execution.

Is a non-compete clause in an AI-drafted employment agreement enforceable?

Post-termination non-competes are generally unenforceable under Section 27 of the Indian Contract Act, 1872. A well-configured drafting tool for Indian startups should flag this and default to confidentiality and non-solicitation clauses instead.

Do ESOP grant letters generated by AI require board approval?

The grant letters themselves do not require board approval, but each grant cycle typically requires a board resolution authorising the grants. The ESOP scheme document must already have shareholder approval in place.

How does FEMA compliance apply to SHA drafting with foreign investors?

FEMA 1999 and the Non-Debt Instruments Rules, 2019 govern pricing, reporting timelines, and eligible instrument types for foreign equity. SHA provisions dealing with buybacks, put options, or assured returns can create FEMA issues and should be reviewed against current RBI master directions.

AI drafting is most effective when embedded in a document workflow, not used ad hoc. In-house counsel who build structured template libraries and compliance checkpoints around AI drafting tools get consistent, reviewable output that holds up under due diligence and litigation scrutiny. Explore how this workflow is supported at legalink.co.in.

Statute citations in this article are verified against LegalInk's verified law registry before publication. Always confirm the current text of any provision before relying on it.

Related posts

← Back to all posts