L1
·
Quiz
·
Lab
L2
·
Quiz
·
Lab
L3
·
Quiz
·
Lab
L4
·
Quiz
·
Lab
Module Test
Module 7 · Lesson 1

The Anatomy of an AI Term Sheet

Why the standard VC template doesn't quite fit when compute costs are a line item and model weights are the crown jewel.
What makes an AI venture term sheet structurally different from a standard SaaS deal — and where do founders lose value without realizing it?

In August 2023, Microsoft's $13 billion commitment to OpenAI and the subsequent $650 million acqui-hire of most of Inflection AI's team by Microsoft in March 2024 put a spotlight on just how differently AI ventures are structured. Inflection had raised $1.3 billion at a $4 billion valuation in June 2023, yet within nine months the majority of its core team — including CEO Mustafa Suleyman — departed under terms that regulators in the UK and EU scrutinized as a de facto acquisition engineered to avoid merger review. The term sheet provisions that governed investor rights, team retention, and IP ownership were at the center of that scrutiny.

Why AI Term Sheets Diverge From the Standard

A conventional venture term sheet concerns itself with three core economic questions: how much equity does the investor receive, what rights protect that equity (liquidation preference, anti-dilution), and what governance rights come alongside it (board seats, protective provisions). For a SaaS company the primary asset is recurring revenue; for a marketplace it is network effects; for a consumer app it is user data. The term sheet is designed to protect those assets.

For an AI venture the primary asset is often a combination of trained model weights, proprietary datasets, compute infrastructure access, and the researchers who built them. None of these map cleanly onto standard IP assignment clauses, and their valuation is radically harder — model weights have no GAAP book value, and a team of five researchers can represent $500 million in competitive moat or nothing at all depending on what they build next.

This asymmetry creates specific pressure points in AI term sheets that standard templates do not address. Founders who sign unmodified NVCA model documents often discover — at Series B or during an acqui-hire conversation — that they have inadvertently signed away rights they assumed they retained.

The Four Structural Differences
Difference 01
IP & Model Ownership
Standard IP assignment clauses cover "inventions and works." Model weights trained on proprietary data may or may not qualify. Ambiguity here has caused disputes at Stability AI (investor suits, 2023–2024) over who owned the core Stable Diffusion weights as the company's finances deteriorated.
Difference 02
Compute Commitments
AI companies often have multi-year GPU or cloud contracts that represent large liabilities. Term sheets increasingly include provisions specifying whether compute commitments are included in "material contracts" requiring investor consent — relevant because a hyperscaler can effectively hold a company hostage via pricing.
Difference 03
Key-Person & Team Retention
OpenAI's board crisis in November 2023 — when Sam Altman was briefly fired — exposed how key-person provisions interact with governance rights. Many AI term sheets now include "key scientist" provisions analogous to key-man clauses in PE, triggering investor put rights if specified researchers depart.
Difference 04
Safety & Use Restrictions
Andreessen Horowitz, Sequoia, and others have begun inserting "responsible AI" covenants as affirmative obligations — requiring AI safety evaluations before model releases. These are governance provisions with material business impact that have no analog in standard term sheets.
Core Term Sheet Vocabulary

Before going deeper into AI-specific provisions, the foundational terms must be precise. Founders frequently confuse pre-money and post-money mechanics, and that confusion has measurable dollar consequences.

Pre-Money ValuationThe agreed company value before the new investment is added. If a VC invests $10M at a $40M pre-money valuation, the post-money is $50M and the investor owns 20% ($10M / $50M).
Liquidation PreferenceThe contractual right of preferred shareholders to receive their investment back (and sometimes a multiple) before common shareholders receive anything in a liquidation, sale, or wind-down. In AI deals, this is especially consequential given acqui-hire structures.
Participating PreferredPreferred shares that first receive their liquidation preference and then participate pro-rata alongside common in remaining proceeds. "Double-dipping." Most AI megadeals use non-participating preferred, but seed rounds often do not.
Pro-Rata RightsAn investor's right to maintain their ownership percentage in future rounds by investing their proportional share. In hyper-competitive AI rounds, fights over pro-rata rights have become a primary negotiating battleground — Microsoft's OpenAI deal specifically structured around these rights.
Information RightsContractual rights to receive financial statements, board observer access, and other data. AI investors increasingly negotiate for model performance metrics and safety evaluation reports as part of information rights packages.
Real Event — Stability AI Investor Dispute, 2023–2024

Coatue Management and other Stability AI investors filed suit in early 2024 alleging that CEO Emad Mostaque had misrepresented the company's financial condition and GPU contract obligations. Central to the dispute were information rights — investors claimed they were not receiving the financial disclosures the term sheet required. The case illustrates how AI-specific liabilities (compute contracts worth hundreds of millions) can make standard information rights provisions inadequate if they don't specifically require disclosure of infrastructure commitments.

The Standard Term Sheet Structure

A term sheet is typically two to five pages and non-binding except for exclusivity (no-shop) and confidentiality clauses. It is organized into three sections: economic terms, control terms, and other terms. Understanding what belongs where is essential before evaluating AI-specific modifications.

SectionKey ProvisionsAI-Specific Pressure Point
Economic TermsValuation, investment amount, option pool, liquidation preference, anti-dilutionOption pool shuffle erodes founder ownership before compute and researcher equity grants
Control TermsBoard composition, protective provisions, voting rights, drag-alongSafety covenants written as protective provisions give investors veto over model releases
Other TermsNo-shop, information rights, pro-rata, ROFR, co-salePro-rata fights in oversubscribed AI rounds; compute contract disclosure in info rights
Founder Insight

The single most expensive line in most AI term sheets is not the liquidation preference — it is the option pool expansion required before closing. A $50M round at a $200M pre-money valuation sounds like 20% dilution. If investors require expanding the option pool from 10% to 20% before closing, the effective pre-money drops to ~$160M, and founders and early investors absorb the full dilution of that pool expansion. This dynamic was documented in detail in Brad Feld and Jason Mendelson's analysis of the "option pool shuffle" — and it hits AI companies particularly hard because researcher comp packages demand large pools.

Reading a Term Sheet as a Negotiating Document

Term sheets are not take-it-or-leave-it documents, even when VCs present them that way. Every provision is negotiable, and the negotiating leverage of an AI founder in 2024–2025 is historically high given competition among investors. The key discipline is identifying which terms are economically material versus which are standard boilerplate that investors expect to keep. Fighting over boilerplate signals inexperience; failing to fight over material terms signals naivety.

In AI specifically, the most material provisions — in rough order of financial consequence — are: (1) liquidation preference structure, (2) option pool size and timing, (3) anti-dilution mechanism (broad-based weighted average vs. full ratchet), (4) pro-rata rights, and (5) IP assignment and model weight ownership. The AI-specific additions — safety covenants, compute disclosure, key-scientist provisions — matter for governance but are rarely the primary value driver.

Lesson 1 Quiz

Term Sheet Anatomy — 3 questions
A VC invests $20M at a $60M pre-money valuation. Before closing, founders are required to expand the option pool from 8% to 18%. What is the effective pre-money valuation that founders should use to calculate their true dilution?
Correct. The option pool shuffle: a 10% expansion before closing on a $60M pre-money means roughly $6M of value transfers from existing shareholders to the pool, making the effective pre-money closer to $48–54M depending on calculation method. Brad Feld's term sheet series documents this exactly.
Not quite. The option pool shuffle is the key concept here — when investors require expanding the pool before closing, that expansion comes out of the pre-money valuation, not post-money. Founders and early investors absorb all the dilution. The effective pre-money is lower than the stated figure.
Which of the following best describes why the Inflection AI / Microsoft situation in March 2024 attracted regulatory scrutiny as a "de facto acquisition"?
Correct. The UK CMA and EU regulators both investigated whether the $650M talent deal — hiring Mustafa Suleyman and most of the Inflection team — constituted a merger that should have been subject to formal review. This is precisely the kind of outcome that term sheet provisions around IP ownership, team retention, and key-person rights are designed to address (or exploit).
Not quite. The regulatory concern centered on Microsoft paying $650M essentially as a licensing/hiring fee to bring Inflection's team and effectively its technology under Microsoft's umbrella without triggering a formal acquisition review — a structure that term sheet provisions on team retention and IP assignment directly enable or constrain.
Which AI-specific term sheet provision has no direct analog in standard SaaS venture deals?
Correct. Safety and responsible AI covenants — requiring specific model evaluation procedures and granting investors veto rights over releases — have no precedent in SaaS or other software deals. They represent a genuinely new category of governance provision specific to frontier AI development, and they carry real business impact since they can delay or block product launches.
Not quite. Liquidation preferences, pro-rata rights, and board composition are entirely standard across all venture deals. The genuinely AI-specific addition is responsible AI / safety covenants that require evaluation procedures before model releases — a governance provision that can materially constrain product timelines in ways that have no SaaS analog.

Lab 1: Term Sheet Anatomy Advisor

AI-assisted practice · Lesson 1

Decode a Hypothetical AI Term Sheet

You are evaluating a Series A term sheet for your AI infrastructure startup. The advisor can walk you through any provision, explain AI-specific implications, and help you identify which terms to prioritize in negotiation. Ask about specific clauses, the option pool shuffle, liquidation preferences, or AI-specific additions you've read about.

Try: "We're being offered $15M at $45M pre-money but they want to expand the option pool from 8% to 20% before closing. How should I think about the true dilution?" — or any term sheet question from Lesson 1.
Term Sheet Advisor
AI Lab
Welcome to the Term Sheet Anatomy lab. I'm your advisor for AI venture term sheet analysis. You can bring me any provision from a real or hypothetical term sheet — I'll help you understand its economic and governance implications, flag AI-specific concerns, and think through negotiating strategy. What would you like to examine first?
Module 7 · Lesson 2

Valuation Mechanics and the AI Premium

How frontier AI companies justify valuations that defy conventional revenue multiples — and when that logic unravels.
When Anthropic raised $4 billion from Google at a reported $18 billion valuation with minimal revenue, what valuation methodology made that number defensible — and what term sheet provisions protected investors if the thesis proved wrong?

In 2023, Anthropic raised two major rounds: a $450 million Series C in May and then secured commitments totaling $4 billion from Google and $1.25 billion from Amazon — part of a broader $7.3 billion raised over the year at valuations peaking around $18 billion. Anthropic's annualized revenue at the time of the Google commitment was estimated at under $100 million. By conventional SaaS multiples of 10–15x revenue, the company was worth $1–1.5 billion, not $18 billion. The gap — roughly 12–18x — represents the AI premium: the valuation that investors assign to competitive positioning, safety research moat, and potential market capture in a winner-take-most dynamic.

But Google and Amazon did not simply write equity checks. Both deals were structured as combinations of equity investment and cloud compute commitments — Amazon's $4 billion included a requirement that Anthropic use AWS infrastructure, and Google's investment was bundled with Google Cloud credits. This structure fundamentally changes the term sheet: compute commitments are both an asset (free infrastructure) and a constraint (vendor lock-in), and the term sheet must account for both.

How AI Valuations Are Actually Calculated

For revenue-generating AI companies, investors use a blended approach combining forward revenue multiples (projecting 24–36 months out), comparable transactions, and qualitative moat assessments. For pre-revenue or early-revenue frontier AI labs, the methodology shifts entirely to option value — the probability-weighted value of capturing a share of an enormous future market.

The standard formulation: if the total addressable market for general AI capabilities is $1 trillion (a commonly cited figure from analysts at Goldman Sachs and Morgan Stanley in 2023–2024), and the company has a plausible 5% probability of capturing 10% of that market, the risk-adjusted value is $5 billion. At a discount rate appropriate for venture-stage risk, this can support a $1–2 billion valuation today. Apply a "strategic scarcity premium" for the small number of credible frontier labs, and the math begins to justify $10–20 billion.

This is not precise finance — it is structured storytelling with numbers. The term sheet provisions that protect investors against this speculation are doing real work.

Anti-Dilution Provisions in AI Mega-Rounds

Anti-dilution protection is the term sheet mechanism that compensates investors if a future round prices lower than their entry (a "down round"). There are three mechanisms, in order of investor friendliness:

Full RatchetThe investor's conversion price adjusts to match the new lower price exactly, regardless of how small the down round is. Extremely investor-friendly, rare in competitive markets, highly punitive to founders. In a down round scenario after an inflated AI valuation, full ratchet can nearly eliminate founder equity.
Broad-Based Weighted AverageThe conversion price is adjusted by a formula that weights both the down-round price and the size of the down round relative to total shares outstanding. This is the market standard for venture deals. Most AI term sheets use this mechanism.
Narrow-Based Weighted AverageSimilar to broad-based but uses a smaller denominator (typically only preferred shares), making it more investor-friendly. Occasionally seen in AI deals where investors negotiate harder given frothy valuations.
Real Event — AI Down Round Risk

Stability AI's valuation fell dramatically from its $1 billion peak (October 2022 funding at that valuation) as the company struggled financially through 2023–2024. While specific down round terms were not publicly disclosed, the company's fundraising difficulties illustrate precisely the scenario anti-dilution provisions address: an AI company raised at a premium valuation based on narrative, then failed to convert that narrative into revenue growth fast enough to sustain the valuation in subsequent rounds. Founders at companies with full-ratchet anti-dilution would face near-total equity wipeout in such a scenario.

Compute-for-Equity Structures

One of the most consequential — and underappreciated — valuation-related provisions in AI term sheets is the compute commitment structure that hyperscalers use when investing. When Microsoft invested in OpenAI, when Google invested in Anthropic, and when Amazon invested in Anthropic, the deals bundled equity investment with cloud compute credits or preferential pricing. This creates a hybrid structure that traditional term sheet analysis does not capture.

From the AI company's perspective: $4 billion in investment that includes $2 billion in AWS credits at market rates is economically equivalent to $2 billion in cash plus a multi-year infrastructure contract. The term sheet must specify: Are compute credits counted as part of the investment for valuation purposes? Do they dilute existing shareholders as much as cash? What happens to the credit obligation if the company changes cloud providers or is acquired?

DealStated AmountStructureKey Term Sheet Issue
Microsoft → OpenAI (2023)$10B (part of cumulative $13B)Cash + Azure compute creditsCompute credits structured as revenue-sharing arrangement; complex liquidation waterfall
Google → Anthropic (2023)$300M–$400M equity trancheEquity + Google Cloud creditsCloud commitment creates vendor dependency; exit rights if cloud SLA not met
Amazon → Anthropic (2023–2024)$4B committedEquity + AWS as primary cloudAWS exclusivity provisions; Trainium/Inferentia chip requirements
Inflection AI (June 2023)$1.3BPure equityStandard preferred; acqui-hire nine months later exposed key-person provision gaps
Liquidation Preferences at Scale

For AI companies raising at high valuations, the liquidation preference stack becomes a critical factor in any M&A outcome. Consider an AI company that has raised three rounds: seed at $10M at $30M pre-money, Series A at $30M at $90M pre-money, Series B at $100M at $300M pre-money. The liquidation preference stack — assuming 1x non-participating preferred throughout — is $140M. This means the first $140M of any acquisition goes to preferred shareholders before common (founders, employees) see anything.

If the company is acquired for $200M — a seemingly positive outcome — founders receive only $60M split among all common holders, while investors received 2.3x their Series B investment and only 1x their earlier rounds. At acquisition prices below $140M, founders and employees receive nothing. This is not a theoretical concern: the acqui-hire market for AI teams frequently produces transaction values in the $100–300M range (talent value, not enterprise value), and the liquidation preference stack determines who captures that value.

Negotiating Leverage Point

The most founder-friendly modification to liquidation preferences in AI deals is a "carve-out" for key employee retention in acqui-hire scenarios — a provision that allocates a specified percentage (typically 5–15%) of any acquisition proceeds directly to the engineering and research team regardless of preference stack order. This provision is increasingly standard in frontier AI term sheets because investors recognize that acquiring parties are primarily buying talent, and talent that receives no acquisition proceeds has incentive to leave immediately post-acquisition. The Google/DeepMind acquisition in 2014 included retention provisions of this type.

Lesson 2 Quiz

Valuation Mechanics — 3 questions
Anthropic's $18B valuation in 2023 was achieved despite revenue under $100M, implying a 180x+ revenue multiple. Which valuation methodology primarily justified this figure?
Correct. For pre-scale frontier AI labs, traditional revenue multiples are inapplicable. The methodology is option value — probability-weighted expected value of capturing a share of an enormous potential market. Strategic scarcity (few credible competitors), safety research moat, and hyperscaler competition for investment further compress valuation discipline.
Not quite. Anthropic's valuation cannot be justified by DCF, comps, or asset values at that stage. The methodology used is option value — the probability-weighted expected value of capturing a significant share of what investors project as a multi-trillion dollar AI market. Strategic scarcity among frontier labs amplifies this further.
An AI company raises at $500M valuation. The next round is a down round at $300M. Under full-ratchet anti-dilution, what happens to the Series A investor's conversion price?
Correct. Full ratchet is the most aggressive anti-dilution mechanism: the conversion price drops entirely to the new down-round price, as if the investor had always been in at that lower valuation. Even a tiny down round (e.g., one share at a lower price) triggers the full adjustment. This is why full ratchet is extremely rare in competitive markets — it can devastate founder and employee equity in any down scenario.
Not quite. The weighted average description actually describes the broad-based weighted average mechanism — the market standard. Full ratchet is more aggressive: the conversion price drops entirely to the new lower price regardless of round size. The fourth option describes broad-based weighted average anti-dilution, not full ratchet.
Why do compute-for-equity structures (like Amazon's $4B Anthropic deal) create term sheet complications that pure cash investments do not?
Correct. The hybrid structure of compute-for-equity deals creates three distinct complications: (1) valuation ambiguity — is $2B in credits worth $2B in economic value? (2) vendor lock-in — the company's technical infrastructure becomes entangled with the investor's commercial interests; and (3) M&A complexity — what happens to committed cloud credits if the company is acquired by a competitor? All of these require explicit term sheet provisions.
Not quite. The core complications in compute-for-equity structures are economic (are credits equivalent to cash for valuation?), strategic (vendor lock-in changes competitive dynamics), and legal (M&A treatment of credit obligations). These require term sheet provisions that standard cash-only deals simply don't need.

Lab 2: AI Valuation Negotiator

AI-assisted practice · Lesson 2

Stress-Test Your Valuation and Anti-Dilution Terms

The advisor can help you model valuation scenarios, evaluate anti-dilution mechanisms, and think through the economic consequences of compute-for-equity structures. Bring hypothetical deal parameters or questions about Anthropic, OpenAI, or Stability AI's funding structures.

Try: "I'm raising at a $200M valuation but my only revenue is $3M ARR. My lead investor is asking for broad-based weighted average anti-dilution. Should I push back on any part of that, and what scenario makes this provision bite me?" — or model any scenario from Lesson 2.
Valuation Mechanics Advisor
AI Lab
Ready to work through valuation mechanics and anti-dilution scenarios. You can give me specific numbers — round size, pre-money, current ARR, existing preference stack — and I'll help you model outcomes under different scenarios. What are we analyzing?
Module 7 · Lesson 3

Control Provisions and Governance Rights

Board composition, protective provisions, and the governance battles that define whether founders or investors control an AI company's direction.
When OpenAI's board fired Sam Altman in November 2023, what governance structure made that possible — and what would standard VC term sheet provisions have done differently?

On November 17, 2023, OpenAI's six-person board fired CEO Sam Altman, citing a loss of confidence in his candor. Within five days Altman was reinstated, three board members resigned, and the governance structure was reformed. The episode is the most scrutinized AI governance event in the industry's history — and it occurred precisely because OpenAI's structure deviates fundamentally from standard VC-backed governance.

OpenAI is a capped-profit LLC controlled by a nonprofit board, not a standard Delaware C-Corp with investor-controlled preferred shares. Microsoft, despite committing $13 billion, had no board seat and no standard protective provisions. The board that fired Altman was selected for mission representation, not capital representation. This structure — which OpenAI chose deliberately to prevent capital from controlling AI development decisions — meant that when the board acted, no investor had a legal mechanism to stop them.

Board Composition: The Power Architecture

In a standard VC-backed company, board composition is negotiated explicitly in the term sheet. The typical progression: at seed stage, founders hold majority control (2-1 or 3-1 founder to investor seats). At Series A, the lead investor takes one seat, creating a 2-1 or 2-2 structure with an independent tiebreaker. At Series B, investors often seek equal representation (2-2-1 or 3-3-1 with independent chair). By Series C, investors frequently have board majority.

For AI companies, this trajectory carries specific risks. An AI company's decisions about which capabilities to develop, what safety standards to apply, and when to release models are both technical and ethical decisions where investor-majority boards may prioritize commercial velocity over risk management. This tension is not hypothetical — it is the explicit reason OpenAI chose its non-standard structure, and it is why Anthropic was structured as a public benefit corporation.

StageTypical Board StructureAI-Specific ConcernFounder Protective Approach
Seed3 founders, 0–1 investor observersFew; founders control all decisionsAvoid giving full board seats at seed
Series A2 founders, 1 investor, 1 independentFirst external check on model release decisionsNegotiate independent seat appointment rights
Series B2 founders, 2 investors, 1 independentInvestor parity; safety covenants get teethEnsure founder protective provisions survive
Series C+Negotiated; often investor-majorityCommercial pressure may override safety judgmentDual-class shares; mission-lock provisions
Protective Provisions: The Investor Veto List

Protective provisions are a list of actions that require preferred shareholder approval (by vote or consent) regardless of board composition. They are the primary mechanism through which investors exercise governance rights without needing a board majority. Standard protective provisions include: issuing new shares, amending the certificate of incorporation, selling the company, declaring dividends, and taking on debt above a threshold.

AI-specific protective provisions added since 2022 have included requirements for: investor consent before releasing models above a specified capability threshold; investor consent before entering data licensing agreements covering training data; and investor consent before making regulatory submissions regarding AI safety classification. These provisions give investors genuine power over core technical decisions.

Real Event — Protective Provisions in Practice

In 2023, multiple AI companies including Cohere and Mistral reported that investors were negotiating protective provisions requiring board notification (and in some cases consent) before entering enterprise agreements with defense or intelligence agencies. This reflects a pattern documented by Bloomberg Law in mid-2023: investors concerned about regulatory risk and reputational exposure are using protective provisions to establish an approval layer over commercial decisions that would be entirely unremarkable in a SaaS company.

Drag-Along Rights and Forced Exits

A drag-along provision allows a specified majority of shareholders (often a combination of preferred majority and common majority) to force all other shareholders to approve a sale of the company. This is designed to prevent a small minority from blocking an otherwise-approved transaction. In AI companies, drag-along provisions interact in complex ways with mission-preservation goals.

Consider: an AI safety-focused company receives an acquisition offer from a company with a significantly different approach to AI development. The board wants to accept; a group of early employees holding common stock wants to block the sale on mission grounds. A well-drafted drag-along provision would override the employee objections. Poorly drafted drag-along provisions — those that don't specify required thresholds clearly — have been litigated in Delaware courts in non-AI contexts and the same vulnerability exists in AI deals.

Drag-AlongAllows a qualified majority of shareholders to compel all others to vote in favor of an approved sale. Critical to specify what majority is required — typically both a majority of preferred and a majority of common voting separately.
ROFR (Right of First Refusal)The company or investors have the right to purchase shares before an existing shareholder sells to a third party. In AI companies with high employee equity, ROFR provisions can prevent secondary market transactions that would reveal sensitive cap table information.
Co-Sale RightsIf a founder sells shares, investors can "tag along" and sell proportionally. Protects investors from founders cashing out while leaving investors in. Particularly relevant for AI founders who may receive acquisition interest for their team specifically.
No-Shop ClauseThe binding provision in an otherwise non-binding term sheet, preventing founders from shopping the term sheet to other investors for 30–60 days. Timing matters enormously in competitive AI rounds where multiple term sheets may arrive simultaneously.
Dual-Class Structures in AI Companies

Several prominent AI companies have adopted or considered dual-class share structures — where founders hold high-vote shares (typically 10:1 or 20:1 voting ratio) that preserve control regardless of equity dilution. Google (2004 IPO), Facebook/Meta (2012 IPO), and Snap (2017 IPO) used this mechanism. In the AI context, Elon Musk's xAI and several private AI infrastructure companies have adopted dual-class structures in their early-stage term sheets.

The investor negotiating tension is significant: sophisticated institutional investors often prohibit or penalize dual-class structures because they eliminate the governance rights that make equity investment valuable. The Institutional Shareholder Services and Glass Lewis both recommend against dual-class structures at IPO. Yet for AI founders who believe mission-critical decisions must not be subject to commercial investor override, dual-class structures may be the only mechanism that reliably preserves founder control through late-stage funding and eventual public market listing.

The OpenAI Lesson for Founders

The OpenAI board crisis did not destroy the company — Altman was reinstated within five days and Microsoft emerged with an observer seat. But it revealed a structural fragility: a governance architecture designed to prevent commercial capture of AI decision-making can also prevent the efficient resolution of leadership conflicts. For founders designing governance: the goal is not to maximize control or maximize investor protection, but to create a governance structure that can make fast, legitimate decisions under pressure. That requires explicit term sheet provisions about board decision thresholds, emergency governance procedures, and what happens when the board is deadlocked.

Lesson 3 Quiz

Control Provisions — 3 questions
When OpenAI's board fired Sam Altman in November 2023, Microsoft — despite committing $13 billion — had no legal mechanism to override the board decision. Why?
Correct. OpenAI's capped-profit structure was explicitly designed to prevent capital from controlling AI development decisions. The nonprofit board governs the entity; Microsoft's commercial investment gives it economic rights in the capped-profit subsidiary but not governance rights over the board. This is why standard VC protective provisions and board rights — which assume a C-Corp structure — simply didn't apply.
Not quite. The key is OpenAI's non-standard structure. As a capped-profit LLC under a nonprofit parent, the governance architecture is fundamentally different from a Delaware C-Corp. Microsoft's position as investor did not translate into board representation or standard protective provisions because the structure was deliberately designed to make that impossible.
A drag-along provision requires "a majority of preferred shareholders" to approve before the drag can be exercised. An AI company has three institutional investors holding 55% of preferred. Two want to sell to a large tech company; one opposes the sale on mission grounds. Can the drag-along be exercised?
Correct. If the drag-along specifies "majority of preferred" and two investors holding more than 50% of preferred approve, the provision can be exercised over any dissenting preferred holders. Well-drafted drag-along provisions typically require both a preferred majority and a common majority (voting separately) to protect against majority-preferred holders forcing sales against the founder and employee interest.
Not quite. A drag-along provision requiring "majority of preferred" can be exercised by holders of more than 50% of preferred shares — no unanimity is required. The scenario describes exactly the condition that triggers the drag-along. The best practice is to require both preferred majority AND common majority voting separately, which would give founders and employees a meaningful check on the sale.
An AI startup negotiating a Series A term sheet is offered the choice: (A) standard board with 2 founders, 1 investor, 1 independent, with investor protective provisions including model release consent; or (B) founder-controlled board with 3 founders, 1 investor observer, no protective provisions. Which consideration most favors Option A from the founder's perspective?
Correct. Counterintuitively, having a well-structured board with independent voices can benefit founders: it adds credibility for future rounds, provides access to investor networks, and creates a governance structure that can make legitimate decisions under pressure. The OpenAI crisis illustrated the downside of governance structures not designed for conflict resolution. Option A's weakness is the protective provisions — founders should negotiate carefully about what specifically requires investor consent.
Not quite. The most nuanced consideration is that structured governance with independent voices often serves founders better than unchecked control — future investors, partners, and regulators all assess governance quality. Option B's appeal (maximum control) must be weighed against its cost (reduced credibility, no external check on decision-making). Option A's protective provisions are the real negotiating point, not the board structure itself.

Lab 3: Governance Architect

AI-assisted practice · Lesson 3

Design Board Structure and Negotiate Protective Provisions

Work through governance design scenarios for your AI company. The advisor can help you evaluate board composition options, draft protective provision language, analyze drag-along risks, and think through how OpenAI's governance lessons apply to your specific situation.

Try: "My Series A lead is asking for a protective provision that requires their consent before we release any model with capabilities above GPT-3.5 level. How should I evaluate this provision and what alternatives exist?" — or any governance question from Lesson 3.
Governance Architecture Advisor
AI Lab
Let's work through your AI company's governance structure. I can help you think through board composition trade-offs, evaluate specific protective provision language, model drag-along scenarios, or analyze how the OpenAI and Anthropic governance structures apply to your situation. What governance question are you working through?
Module 7 · Lesson 4

Negotiating Strategy and Closing Dynamics

How to run a competitive process, what to fight for, what to concede, and how AI-specific provisions have shifted the negotiating table since 2022.
Given that AI founders in 2023–2024 often received multiple term sheets simultaneously from top-tier investors, what negotiating strategies maximized founder outcomes — and which backfired?

Mistral AI was founded in April 2023 by former DeepMind and Meta researchers. By June 2023 — just weeks after founding — the company had raised a €105 million ($113M) seed round at a €240 million pre-money valuation, with investors including Lightspeed, Index Ventures, and Xavier Niel. The round was oversubscribed with multiple competing term sheets. In December 2023, Mistral raised an additional €385 million Series A at a €2 billion valuation — approximately 8x the seed valuation in six months, with no shipped product and minimal revenue.

The negotiating dynamics in both rounds were unusual: Mistral's founders — having deep credibility from Llama and Gemini work — were able to run a competitive auction process that generated multiple competing bids, drove valuation significantly above initial offers, and resulted in founder-favorable governance terms including no safety-covenant protective provisions in the standard form. This outcome was not accidental — it was the product of deliberate process management.

Running a Competitive Process

The fundamental principle of venture fundraising negotiation is that leverage comes from alternatives. A founder negotiating with a single interested investor has almost no leverage. A founder with three competing term sheets has significant leverage on every economic and governance provision. Building that competitive dynamic is the primary goal of a fundraising process, and it is substantially more achievable for AI companies in 2024 than for almost any other category of startup.

The mechanics of a competitive process: (1) identify and approach all plausible investors simultaneously, not sequentially; (2) create artificial time pressure by setting a decision deadline that aligns with multiple investors' internal processes; (3) use anchoring — share the highest offer (or a credible number) with other interested investors to pull their bids upward; (4) leverage strategic investor interest (hyperscalers) to pressure pure financial investors, and vice versa.

What to Negotiate: The Priority Stack

Founders often spend negotiating capital on visible but economically minor issues (valuation) while ceding invisible but material provisions (option pool size, liquidation preference structure, anti-dilution mechanism). The effective negotiating approach priorities provisions by their expected financial impact across likely exit scenarios.

ProvisionFight Hard?Expected Financial ImpactAI-Specific Note
Option Pool Size / TimingYes — highest priorityDirect equity dilution, often 5–15% of founder ownershipAI researcher comp requires large pools; fight pre vs. post-money expansion
Liquidation Preference StructureYes — 1x non-participating is the lineCan determine whether founders receive anything in an acqui-hireAcqui-hire is the modal AI exit; preference structure determines outcome
Anti-Dilution MechanismYes — broad-based weighted average onlyPotentially total equity wipeout in down roundHigh AI valuations make down round risk real; full ratchet unacceptable
Board CompositionNegotiate carefully, not aggressivelyModerate — governance control has indirect economic valueIndependent seat appointment is the key leverage point
Pro-Rata RightsGrant standard, fight super pro-rataModerate — determines future round dynamicsSuper pro-rata in AI rounds can lock out better investors in later rounds
Safety CovenantsNegotiate scope carefullyLow economic impact, high operational impactDefine "material release" precisely; broad language creates operational risk
Information RightsGrant standard; add compute disclosureLow direct economic impactEnsure compute contract disclosure is included; investor suit risk if omitted
ValuationImportant but often overweightedHigh visibility, moderate actual impact if other terms are rightInflated AI valuations create down-round risk — be careful what you win
The No-Shop and Timing Dynamics

The no-shop clause — binding even in an otherwise non-binding term sheet — prevents founders from soliciting competing offers for a specified period (typically 30–60 days). In AI markets where deal velocity is high and investor interest is intense, the no-shop period is a significant concession. Once signed, founders lose the competitive leverage that generated the term sheet in the first place.

Negotiating strategy on no-shop: (1) shorten the period to 30 days maximum; (2) add a right to continue discussions with investors already in conversation before signing; (3) add a "fiduciary out" allowing termination if a substantially better offer arrives; (4) add a reciprocal exclusivity — the investor agrees not to invest in directly competing companies during the no-shop period. The fourth provision is rarely granted but signals negotiating sophistication.

Real Event — Competitive AI Round Dynamics, 2023

According to reporting by The Information and Bloomberg in 2023, multiple AI companies including Character AI, Cohere, and Adept AI received simultaneous term sheets from competing investors including Andreessen Horowitz, Sequoia, Khosla Ventures, and Google Ventures. In several cases, the competitive dynamic drove valuations 40–60% above initial offers. Character AI's August 2023 round ($150M at $1B valuation from a16z) followed a competitive process in which Google reportedly offered a lower valuation. The tactical lesson: creating genuine competition among investors is worth more than any single negotiating tactic on individual provisions.

AI-Specific Closing Risks

The period between term sheet signing and closing — typically 30–90 days for a venture round — is particularly fraught for AI companies because of factors that don't apply to most other sectors. Regulatory review risk: several AI deals in 2023–2024 attracted antitrust scrutiny (Microsoft/OpenAI, Google/Anthropic) that delayed or complicated closing. Key-person departure risk: AI research teams are highly mobile, and a key researcher departure between term sheet and close can trigger investor renegotiation or withdrawal. Model capability announcements: a competitor's major model release between term sheet and close can reset market valuation expectations and trigger investor attempts to renegotiate.

Term sheet provisions addressing closing risk: (1) a "bring-down" certification from founders at closing confirming no material adverse change; (2) a material adverse change (MAC) clause that specifies what events give investors the right to renegotiate or walk away; (3) an explicit definition of "material adverse change" that excludes market and industry conditions (standard in M&A but less common in early-stage term sheets). AI-specific MAC definitions should address team departures, regulatory actions, and compute access disruption.

Material Adverse Change (MAC)A provision allowing investors to renegotiate or exit if a defined set of significant negative events occurs between term sheet signing and closing. AI-specific MACs often include: key researcher departure, regulatory action against AI activities, loss of compute access, and major competitive capability release.
Bring-Down CertificateA representation at closing that all representations made at term sheet signing remain true. Founders sign this at closing; material misrepresentations create legal liability. Particularly important in AI for compute contract obligations and team composition representations.
Super Pro-RataAn investor's right to invest more than their proportional share in future rounds — e.g., the right to invest up to 3x their ownership percentage. In AI rounds, super pro-rata rights can lock out better investors and give one early investor disproportionate influence over future rounds.
Post-Signing: What Founders Get Wrong

The term sheet is the beginning of an investor relationship, not the end of negotiation. The definitive documents — stock purchase agreement, investor rights agreement, voting agreement, right of first refusal agreement — are where the term sheet provisions are converted into binding legal language. This conversion introduces ambiguity that founders often allow investors' lawyers to resolve in investors' favor.

The three most common post-signing errors: (1) accepting investor-drafted definitive documents without parallel founder counsel review; (2) allowing vague term sheet language to be resolved toward investor interpretation in definitive docs (e.g., "reasonable best efforts" vs. "best efforts" in safety covenants); (3) failing to negotiate the sequence of closings in tranched rounds — an investor who commits $20M in two $10M tranches holds leverage over the second tranche that can be used to renegotiate terms after you've spent the first tranche.

Closing Principle

The best documented advice on AI venture term sheet negotiation comes from founders who have done multiple rounds: the most expensive mistake is optimizing for valuation at the expense of structure. A $500M valuation with 2x participating preferred and a large option pool requirement is economically worse for founders than a $400M valuation with 1x non-participating preferred and a reasonable pool. Investors know this arithmetic; many founders, especially first-time founders in their first institutional round, do not. The purpose of this module is to ensure you are not in that category.

Lesson 4 Quiz

Negotiating Strategy — 3 questions
Mistral AI raised €105M in June 2023 just weeks after founding. Which factor most directly explains the favorable terms founders received in that round?
Correct. Mistral's founders had elite credibility (Llama and Gemini research backgrounds), and they ran a competitive simultaneous process. Both factors were essential: credibility generated investor interest, and competitive process generated leverage. Neither alone is sufficient — elite credentials without a competitive process leaves money and terms on the table, while a competitive process without a credible team fails to attract quality investors at all.
Not quite. Mistral had no product and no revenue at the time of the seed round. The favorable terms resulted from the founders' research credibility (generating genuine investor competition) combined with deliberate process management — running simultaneous conversations with multiple investors to create a competitive dynamic. That combination of credential and process is what produced the €240M pre-money seed valuation.
An AI company signs a term sheet with a 45-day no-shop clause. On day 30, an unsolicited offer arrives that is 50% higher in valuation with better terms. Which no-shop provision would have given the founders the most flexibility to respond?
Correct. A shorter no-shop (30 days rather than 45–60) combined with a fiduciary out is the founder-optimal structure. The fiduciary out — allowing the board to engage with materially superior unsolicited offers — is standard in public M&A but less commonly negotiated in private venture term sheets. Combined with a short period, it preserves flexibility for exactly the scenario described. A right of first refusal for the original investor is far weaker protection — the competing investor may not want their offer shopped to the original investor.
Not quite. The founder-optimal no-shop structure is (1) shortest possible period and (2) a fiduciary out allowing engagement with materially superior unsolicited offers. Longer periods and unlimited no-shops maximize investor protection at founders' expense. The ROFR option is worse than a fiduciary out because it discourages competing investors who don't want their offer used as a price floor for the original investor to match.
An investor offers a $300M valuation with 2x participating preferred, versus a competing offer of $250M with 1x non-participating preferred. Both invest $20M. The company is later acquired for $80M. Which deal results in more proceeds to founders and employees (common shareholders)?
Correct. This is the core lesson of preference structure analysis. Under Deal B: investor gets $20M (1x preference), common holders split $60M. Under Deal A: investor gets $40M (2x preference), then participates pro-rata in the remaining $40M — if investors own ~6.25% ($20M of $320M post-money), they receive an additional ~$2.5M, leaving ~$37.5M for common. Deal B produces roughly $60M for common; Deal A produces roughly $37.5M. A $50M valuation "win" is worth negative $22.5M in this exit scenario.
Not quite. Work through the math: Deal A (2x participating): investors get $40M first, then ~6.25% of remaining $40M = ~$2.5M more; common holders receive ~$37.5M. Deal B (1x non-participating): investors get $20M; common holders receive $60M. Deal B is substantially better for common shareholders despite the lower valuation. This is why experienced founders prioritize preference structure over headline valuation.

Lab 4: Term Sheet Negotiation Simulator

AI-assisted practice · Lesson 4

Run Your Negotiation Strategy

The advisor plays the role of a strategic negotiation coach. Describe your deal situation — multiple term sheets, a specific investor's demands, a preference structure dispute — and work through the negotiating strategy, priority stack, and likely outcomes. You can also model acquisition scenarios to test whether specific term combinations work in your favor.

Try: "I have two term sheets: Sequoia at $180M pre-money with 1x non-participating preferred, and a strategic investor at $220M pre-money with 1x participating preferred and a safety covenant requiring their consent before any model release. How do I think about which one to take and what to counter-offer?" — or bring any real negotiating scenario from Lesson 4.
Negotiation Strategy Coach
AI Lab
Welcome to the negotiation strategy lab. I'm your coach for AI venture term sheet negotiations. Bring me your deal situation — term sheet details, competing offers, specific provisions you're uncertain about — and we'll work through the priority stack, model economic outcomes across exit scenarios, and develop your counter-offer strategy. What deal are we working on?

Module 7 Test

Term Sheet Dynamics for AI Deals · 15 questions · Pass at 80%
1. Which of the following is the most AI-specific addition to a standard venture term sheet?
Correct. Safety covenants with investor veto rights over model releases have no precedent in non-AI venture deals.
The correct answer is safety/responsible AI covenants — the only provision listed with no SaaS or standard venture analog.
2. The "option pool shuffle" refers to which practice?
Correct. The option pool shuffle is a pre-closing expansion that reduces the effective pre-money valuation founders receive.
The option pool shuffle is the pre-closing pool expansion requirement that dilutes existing shareholders, effectively reducing the founders' true pre-money valuation.
3. Inflection AI raised $1.3 billion in June 2023 at a $4 billion valuation. By March 2024, most of its team had joined Microsoft for approximately $650 million. Regulators investigated this as a potential de facto acquisition primarily because of concerns about:
Correct. The UK CMA and EU regulators examined whether the structure evaded merger control — effective acquisition of the team and technology without filing for antitrust approval.
The correct answer relates to regulatory evasion — the talent/licensing deal effectively acquired the company's assets without triggering formal merger review requirements.
4. Under full-ratchet anti-dilution, a company's down round at 60% of the original valuation causes the prior investor's conversion price to:
Correct. Full ratchet is the most aggressive mechanism: the conversion price drops entirely to the new down-round price regardless of round size.
Full ratchet resets the conversion price completely to the new lower price. The weighted average description applies to broad-based weighted average anti-dilution, not full ratchet.
5. Amazon's $4 billion investment in Anthropic created a term sheet complication absent from standard cash investments because:
Correct. Compute-for-equity structures create three novel complications: valuation ambiguity, vendor dependency, and M&A complexity around credit obligations.
The correct answer is the hybrid cash+compute structure, which creates valuation ambiguity, vendor lock-in, and M&A complications that pure cash investments don't involve.
6. OpenAI's governance crisis in November 2023 revealed that Microsoft, despite $13 billion in commitments, had no mechanism to override the board's decision to fire Sam Altman. The structural reason was:
Correct. The nonprofit-controlled capped-profit structure is why standard VC governance rights — board seats, protective provisions — didn't exist in the expected form for Microsoft.
The correct answer is OpenAI's unique non-standard structure: a nonprofit-controlled capped-profit LLC that wasn't subject to normal C-Corp investor governance mechanics.
7. A drag-along provision that requires only "a majority of preferred shareholders" without requiring common shareholder approval creates which founder risk?
Correct. A preferred-only drag-along can force founders and employees (common shareholders) into a sale they oppose, with no common vote required.
The correct risk: without a required common shareholder vote, preferred investors can drag founders and employees into an approved sale regardless of their opposition or economic interest.
8. Which provision in an AI term sheet most directly determines founder proceeds in an acqui-hire exit at $150M where the liquidation preference stack is $120M?
Correct. At $150M with a $120M stack, the participating vs. non-participating distinction determines whether founders receive all $30M remaining or share it pro-rata with participating preferred investors.
The key determinant is participating vs. non-participating preferred. With $30M above the preference stack, non-participating preferred leaves all $30M for common; participating preferred splits that $30M between investors and common pro-rata.
9. Mistral AI's exceptional seed round terms in June 2023 (€105M at €240M pre-money, minimal product) were primarily the result of:
Correct. Credential-driven investor demand plus deliberate competitive process management — not any single structural provision — generated Mistral's market-beating terms.
The correct answer combines two factors: extraordinary founder credibility (generating multiple competing term sheets) and deliberate process management to leverage that competition.
10. A "super pro-rata" right gives an investor the ability to:
Correct. Super pro-rata lets an early investor grow their stake in later rounds beyond what their current ownership would normally entitle them to — potentially crowding out later-stage investors.
Super pro-rata is the right to invest more than proportional ownership in future rounds. It can crowd out better investors and give early backers disproportionate influence over subsequent rounds.
11. The Stability AI investor dispute of 2023–2024 (Coatue Management suit) primarily centered on which term sheet provision?
Correct. The Coatue suit alleged failures of the information rights obligations — specifically inadequate disclosure of compute contract liabilities and financial condition.
The correct answer is information rights — Coatue alleged inadequate disclosure of financial condition and compute contract obligations that the term sheet required the company to provide.
12. A broad-based weighted average anti-dilution mechanism differs from narrow-based weighted average by:
Correct. The "broad base" refers to using all shares (common + all preferred) as the denominator — the larger denominator produces a smaller, less punitive conversion price adjustment. Narrow-based uses only preferred, producing a more investor-favorable (more founder-punitive) result.
The broad vs. narrow distinction is about the denominator in the weighted average formula. Broad-based includes all shares outstanding (giving a larger denominator and smaller adjustment); narrow-based includes only preferred (giving a smaller denominator and larger, more investor-favorable adjustment).
13. A Material Adverse Change (MAC) clause in an AI venture term sheet that is drafted to include "competitive landscape changes" as a triggering event would be:
Correct. Including competitive changes as MAC triggers is extremely dangerous for AI founders given the pace of model releases. A GPT-4 or Claude 3 release between signing and closing could be characterized as a competitive MAC, giving the investor leverage to renegotiate or exit.
This provision is highly problematic for founders. In a rapidly evolving AI market, including competitive changes as MAC triggers gives investors a perpetual renegotiation option — any major model release by a competitor could qualify. Founders should explicitly exclude industry and market conditions from MAC definitions.
14. Anthropic's public benefit corporation structure, like OpenAI's capped-profit structure, was adopted primarily to:
Correct. Both Anthropic's PBC and OpenAI's capped-profit structures were designed to ensure that safety and mission considerations have legal standing in governance — not merely ethical standing — against purely commercial investor pressure.
The correct answer is mission protection: these structures give safety and mission considerations legal standing in governance, ensuring that fiduciary duty includes mission and not just return maximization for shareholders.
15. The single highest-priority term sheet provision for founders in AI acqui-hire exit scenarios — where acquisition proceeds are often in the $100–300M range — is:
Correct. In acqui-hire scenarios with $100–300M proceeds, a large preference stack (multiple rounds of 1x or higher preference) can absorb nearly all proceeds before common shareholders see any returns. Liquidation preference structure is the highest-stakes provision for this exit type.
The correct answer is liquidation preference structure. In acqui-hire transactions — the most common AI exit type at current market prices — the preference stack often equals or exceeds the acquisition price, meaning founders and employees receive nothing. 1x non-participating preferred, negotiated on every round, is the primary protection.