Does government equity dilute your fusion job offer?

What did fusion startups just hand the government?

In late October 2026, Helion and Commonwealth Fusion Systems accepted $260 million in federal semiconductor awards and agreed to put the US government on their cap tables as a non-voting shareholder. Helion took $175 million and CFS $85 million, both from the Department of Commerce under the CHIPS and Science Act, earmarked for the advanced chips that power and control their reactors. Helion priced the government's shares off its Series G, which valued the company at $15.5 billion — so roughly a 1% slice. CFS hasn't disclosed its latest valuation, so nobody outside the company knows the size of its stake. Several other fusion firms, including Thea Energy, Avalanche Energy and Pacific Fusion, say they've given up no equity at all.

Set against what these companies have already raised, the money is modest. CFS has pulled in roughly $4 billion and Helion has about $3.2 billion in committed capital, so $260 million buys perhaps a few months of burn. What's new is the structure. The Fusion Industry Association had lobbied for a single large federal injection into fusion research — on the order of $10 billion, with billions ring-fenced for startups. What landed instead came out of Commerce, with equity strings attached. That pattern now has precedent: the government converted $8.9 billion into roughly 10% of Intel, and took about 10% of USA Rare Earth alongside a $277 million award and a $1.3 billion loan.

Why should a candidate care about a non-voting minority stake? Because the cap table behind your offer letter is the thing that decides whether your options ever turn into money, and this deal tells you two useful things about it. First, federal dollars are now arriving as equity rather than grants, which means more shareholders ahead of you over the next decade. Second, the companies taking the deal are the ones scaling hardest and hiring hardest. If you're interviewing in fusion, batteries, advanced nuclear or chips, you're negotiating against a capital structure that's getting more crowded — and you should price that in rather than ignore it.

Does a government equity stake dilute your stock options?

Yes, but far less than the headline suggests. A newly issued 1% non-voting stake cuts the value of your grant by roughly 1% — not by one percentage point of ownership. If your options represent 0.05% of the company, you end up at about 0.0495%. On a grant that might be worth $200,000 on paper, that's a couple of thousand dollars of theoretical value. And because Helion priced the government's shares off its most recent round rather than issuing them cheaply, existing holders weren't handed a discount at their own expense. Non-voting also means no new party at the table demanding board seats, protective provisions or a veto over the sale that eventually pays you out.

The dilution that actually moves your number sits elsewhere. A fusion plant costs billions, which means several more rounds before anyone sees revenue, and each one issues new preferred shares ahead of your common stock. Each round also tends to refresh the option pool — pools typically run 15–20% of fully diluted shares after Series A, topped up as the company grows, and that top-up dilutes everyone who came before. Then there's the preference stack: every dollar of preferred liquidation preference gets paid before common holders see anything. On a company that has raised $3–4 billion, that stack is enormous, and it's the single most under-asked question in deep-tech offers.

Here's the contrarian part. I'd spend almost no negotiating energy on the government's stake and a lot on terms nobody volunteers. Ask whether the federal shares were issued at the last round's price or below it, whether they carry warrants or anti-dilution protection, and whether any award money is clawed back if milestones slip. A plain non-voting stake priced at fair value is a non-event. A stake with a ratchet, a liquidation preference or a repayment trigger is a real claim on your upside. The same logic applies at AI labs and other richly valued private companies — the discipline of valuing a pre-IPO grant on its actual terms matters more than the valuation in the press release.

Which questions should you ask before signing a fusion offer?

Five questions separate a real equity offer from a decorative one: the fully diluted share count, your strike price and the date of the last 409A valuation, the aggregate liquidation preference in dollars, your post-termination exercise window, and the terms of any government or strategic stake. Ask them in writing, by email, before you accept. Any company that intends to pay you partly in equity should be able to answer all five in a paragraph, and the ones that stall are telling you something. You don't need the full cap table — you need enough numbers to turn "0.05%" into a range of outcomes you can actually compare against a competing offer.

Pay particular attention to the two prices hiding in every grant. Recruiters quote your equity against the preferred price investors paid in the last round, because that's the bigger number. Your strike price is set by the 409A appraisal of common stock, which usually comes in well below it — commonly somewhere between a fifth and half of the preferred price at Series B and later, with the gap narrowing as an exit looks more likely. That spread is where your upside lives, so ask when the 409A was last refreshed. Joining just before a repricing can mean a materially higher strike on identical paper, and nobody will flag it for you.

The exercise window is the term people regret ignoring. The standard is 90 days after you leave, which means that if you quit a pre-revenue fusion company after four years of vesting, you have three months to find real cash for the strike price and, with ISOs, potentially pay alternative minimum tax on a spread you can't sell. Push for two to five years, in writing, in your offer letter. Ask about single- versus double-trigger acceleration too. And when you're mapping offers side by side, request illustrative exit scenarios showing what reaches common holders after debt and preferences — not a percentage multiplied by a headline valuation.

Is the base salary good enough to take less equity?

At the best-funded fusion companies, base pay is now competitive enough that you can reasonably trade equity for cash. Public 2026 job postings put senior and staff engineering roles at Commonwealth Fusion roughly in the $130,000–$205,000 band, with principal-level RF systems work advertised as high as about $260,000. Helion's postings generally ran higher at equivalent levels — a materials engineer range around $159,000–$208,000, and senior plant software engineering in the $240,000–$255,000 region, partly reflecting the Seattle-area market. Across the sector, entry-level engineers tend to land between $75,000 and $110,000, mid-career roles between $110,000 and $155,000, and senior, principal and management roles at $150,000 and up.

Those numbers give you a lever. When the liquidity event is undatable, cash today is worth more than paper tomorrow, and the companies know it. The move that works: name a specific trade rather than asking for "more." Try this — "I'm excited about the role and I believe in the technology. Given that first revenue is several years out, I'd rather be weighted toward cash: can we move $20,000 from the equity grant into base, or keep the grant and add a $25,000 signing bonus?" If base is genuinely capped by a band, ask for a mid-cycle review at six months, an annual refresh grant in writing, or the extended exercise window instead.

Leverage comes from having two live processes, and that starts well before the money conversation. Deep-tech hiring teams screen hard on specifics — the vacuum systems you've commissioned, the cryogenic or high-field magnet work you've touched, the plasma diagnostics you've written code for — and generic engineering CVs get filtered out before a human reads them. If you're not getting to final rounds at more than one company, you're negotiating against yourself. Running your CV through a free AI CV analysis will tell you where the impact and ownership signals are thin and whether the file parses cleanly, which is a cheaper fix than another month of silence.

How do you price equity when the timeline keeps slipping?

Assume the date slips, then discount accordingly. Fusion's published milestones have moved repeatedly: SPARC's first plasma, once targeted for 2025, is now aimed at 2027, with net energy gain to follow "as fast as humanly possible," and CFS's first commercial plant, the 400-megawatt ARC outside Richmond, Virginia, is slated for the early 2030s. Helion demonstrated measurable deuterium-tritium fusion in Polaris in February 2026 at 150 million degrees Celsius and says it's still on schedule for first electrons to the grid in 2028 under its 50-megawatt Microsoft power agreement. For scale on how these schedules behave, ITER pushed its own first plasma from 2025 to 2034.

That has a concrete consequence for your grant: a standard four-year vest will almost certainly complete before the company has revenue, let alone liquidity. A sane way to model it is to haircut the 409A-based value of your equity by something like 25–50% at the later private stages, and more if the company is earlier, then ask what happens at year four. Does the company issue refresh grants? On what cadence, and at what size? Get the answer before you sign, because the moment you're fully vested with no refresh in sight is exactly when a competitor's cash offer starts looking irresistible — and when your negotiating position is weakest.

Then ask about the road to cash. Many large private companies now run periodic employee tender offers, and understanding how a startup tender offer actually works tells you whether your shares have any realistic path to money before an IPO. Ask whether the company has ever run one, whether secondary sales are permitted at all, and who approves transfers — the rules governing early share sales vary wildly by company. A non-voting government shareholder shouldn't block any of that. My recommendation: treat a federal stake as a credibility signal, price your equity as a lottery ticket with a long fuse, and negotiate base pay like the exit is five years later than promised. That's the same discipline that applies when an expected IPO keeps getting pushed back.

Frequently asked questions