In 1937, a pharmaceutical company introduced a liquid version of sulfanilamide, a widely used antibacterial drug. To create the new formulation, its chemist needed a solvent that would dissolve the medicine, flow easily into bottles, and taste reasonably pleasant. He chose diethylene glycol, an industrial solvent that appeared to satisfy every requirement the company had considered.

The problem was that no one had considered whether the solvent itself was safe.

Warning signs about diethylene glycol’s toxicity already existed, but the formulation was never tested for safety before being sold. More than 100 people died.

The disaster became the direct catalyst for the Federal Food, Drug, and Cosmetic Act of 1938, which for the first time required manufacturers to demonstrate drug safety before marketing a product.

The sulfanilamide worked exactly as intended. It was what carried it that caused the damage.

Many founders sign SAFEs the same way. The capital, the round, all work smoothly with little friction. But the risk lurks in the vehicle: conversion mechanics, stacked dilution, and clauses that activate at moments you didn’t anticipate at signing.

This week: How the SAFE closes fast and stays quiet, until it doesn’t, and what you need to do before the math catches up with you.

Founders who raise on SAFEs see the part of the instrument that’s above the waterline. They’re attracted by the brevity, speed to close, the lack of conversion deadlines, interest, and the low legal costs.

But its real power only becomes visible at the moment your startup makes contact with its terms:

Optional clauses pack big, invisible giveaways: The standard SAFE template includes optional provisions that look standard but matter enormously. The Most Favored Nation (MFN) clause, for example, means that every subsequent SAFE issued on better investor terms (a lower cap, for example) automatically upgrades every prior SAFE holder to those same terms, without any new signatures or negotiation. If you issued a SAFE with investor A with an MFN clause, and subsequently closed investor B with a lower cap, you’ve retroactively amended investor A’s instrument without realizing it.

Stacking compounds founder dilution. Post-money SAFEs guarantee ownership to the holder. But when you start stacking post-money SAFEs all future dilution comes out of your stake as the founder. Three post-money SAFEs at 10% each remove 30 percentage points from your founder stake before the priced round, the option pool, or the new investor enters the math.

Brevity creates litigation risk. The brevity of the SAFE creates ambiguity that’s a standing invitation for future litigation. The details vary, but it boils down to parties having different understandings of what the specific clauses meant, triggering the possibility of future lawsuits to resolve them. The standard SAFE is silent on many critical points that have enormous consequences: what happens if a subsequent SAFE is issued at a different cap, or precisely how conversion works when a mix of instruments is outstanding. Lawyers’ fees are the most expensive way to resolve them.

The Unlock

Signing a SAFE locks in three things simultaneously: what you have already given away, what this instrument obligates you to, and what conversion will cost you at your next raise. All three are fixed at signing but only become visible when you can’t renegotiate any of them.

The stack is compounding, and starts when you sign the first instrument. Most founders think about each SAFE in isolation: this round, this investor, this cap. That framing collapses when you’re dealing with post-money SAFEs because each post-money SAFE sits on top of everything already committed. Your baseline ownership is capped at what remains after every prior SAFE has taken its guaranteed percentage.

The most damaging clauses are the most silent at signing. MFN provisions and early-exit terms appear innocuous at signing but create significant problems at acquisition offers and subsequent closings. These are the moments when you as the founder have the least bandwidth and the highest stakes. When you sign an MFN clause and issue a lower-cap SAFE later, you’ve retroactively improved every prior investor’s return at the expense of your own payoff.

The conversion math is a present commitment expressed in future numbers. Every SAFE signed today is a fully binding equity agreement. If you arrive at your Series A term sheet without having run the conversion math, you’ll risk discovering the effects of what you gave away months or years ago, under conditions you might no longer remember, with no ability to change any of it.

From The Capital Playbook

Know your stack position before you sign anything.

Before closing any SAFE, document every outstanding convertible instrument: each SAFE, its cap, whether it is pre-money or post-money, and the ownership percentage it guarantees its holder. That total is what you have already committed. Every new SAFE you sign moves from what remains, not from 100%. If you don’t calculate this before signing, you’re agreeing to give away a percentage of a number you’ve never looked at.

Read every optional clause and identify what triggers it.

Before signing, identify every optional clause in the instrument. If an MFN provision is present, what does it mean in practice? What is the lowest cap you have issued so far, and are you prepared for every prior SAFE holder to receive that cap automatically if you issue better terms to the next investor? If an early-exit provision is present, what is the liquidity event definition and what will each SAFE holder would receive in an acquisition at a typical valuation?

Run the conversation calculation at three scenarios before closing.

What would conversion look like at a conservative, expected, and strong valuation at Series A? For each scenario, what would each SAFE holder’s resulting ownership, and your ownership be after all conversions? Use Carta benchmarks as your base case. If any scenario produces a founder ownership number that would make the company difficult to fund or leave the founding team without sufficient equity to stay motivated, now is the time to fix it.

Look out for what happens between Seed and Series A, not just this round.

The median time from seed to Series A is now 2.2 years, per Carta, which is the longest on record. A SAFE you sign today likely sits on your cap table for the better part of two years while you may issue additional instruments, extend runway on bridge SAFEs, and face an option pool expansion at close. Can you tell what stacking looks like across that entire timeline, not just at the moment of signing?

Proof Points

“Temporary” SAFEs are governing equity decisions that will compound for two to three years.

Carta’s analysis shows two out of every 5 startups now take three or more years to get from seed to Series A, more than double the rate in 2019. Meanwhile, median Series A pre-money valuations hit $49.3M in Q3 2025 and kept climbing. So if you signed a post-money SAFE at a $10–15M cap, you’ve locked in conversion against a moving target that’s working against you.

The ownership trajectory is steeper than most founders anticipate, and it begins before the first priced round.

Carta’s Founder Ownership Report 2026, drawing on data from over 45,000 startups, shows the median founding team retains 56% after seed, 36% after Series A, and 23% after Series B. That 20-percentage-point drop from seed to Series A, which is the single largest step-down in the entire trajectory, is the moment when all outstanding SAFEs convert simultaneously alongside the new investor’s ownership. The 36% post-Series A figure is the median outcome for founders who managed their cap table carefully. The outcomes below it are the result of unmodeled SAFE stacks.

The instrument’s brevity increases, rather than decreases, frequency and cost of disputes.

Bloomberg Law’s December 2025 identifies ambiguous ownership terms and unspecified valuation methodology as the two most consistent triggers of founder-investor disputes, which are the precise conditions a SAFE’s brevity creates. A document that takes thirty minutes to sign can take years and significant legal cost to resolve when two parties discover they understood its terms differently.

The One Thing

Before you sign your next SAFE or before you close the round currently in progress, do this simple exercise:

The stack position: List every SAFE currently outstanding. For each one, note the investment amount, the valuation cap, and whether it is pre-money or post-money. Calculate the total ownership percentage you have already committed. That number is your starting point. Everything you sign from here reduces it further.

The clause inventory: Does this SAFE contain an MFN provision? An early-exit provision? A pro-rata right? For each one present, write one sentence describing the specific condition that triggers it and what you are obligated to do when it does. If you cannot write that sentence, you do not yet understand what you have agreed to.

The conversion table: At a $15M, $30M, and $50M Series A pre-money, what percentage does each SAFE holder receive? What is your ownership after every conversion? What does a 10% option pool expansion do to that number? If any scenario leaves you below the ownership threshold you need to stay motivated and investable through the next five years of building, you need to know that before you sign, not when the term sheet is on the table and the lead investor’s lawyer is waiting for your reaction.

A SAFE takes thirty minutes to sign. These three steps take an afternoon. The difference is where you up your game.

Build wisely!