C3.ai: The Founder's Back, the Losses Are Smaller, and Somehow That's the Good News

# C3.ai: The Founder's Back, the Losses Are Smaller, and Somehow That's the Good News

Here's a fun exercise: try to find two Wall Street price targets on C3.ai that agree with each other. UBS says $16. Consensus says $8. Another shop says $10.09. That's not analyst disagreement — that's analysts describing three different companies. And that, more than anything, tells you where C3.ai sits right now: a stock nobody can quite decide whether to bury or bless.

Let's start with the headline that got everyone excited. Back on June 3, C3.ai reported Q4 FY2026 earnings and beat estimates — a loss of $0.33 a share against expectations of a $0.45 loss. Progress, technically. Losing less money than expected is still losing money, but on Wall Street "less bad" has a way of getting dressed up as "good." Add to that the return of founder Thomas Siebel to the CEO chair, announced in the same breath as those preliminary results back in May, and you've got the makings of a turnaround story: the guy who built the thing coming back to fix the thing.

Except the turnaround story has a revenue problem. A big one. Internal figures show revenue down 52.5% quarter-over-quarter and off 35.67% on a trailing-twelve-month basis. Operating margins sit at a brutal -194.86%. Cash burn hit $190 million in a single quarter. This isn't a company narrowing its losses because the business is healing — it's a company narrowing per-share losses while the top line falls off a cliff, which is a very different, much scarier story once you actually look under the hood.

And then there's the small matter of the founder selling his own stock. Siebel offloaded $1.3 million worth of shares on August 4 — just days ago as I write this, and just weeks after retaking the CEO title. Maybe it's estate planning. Maybe it's a 10b5-1 plan on autopilot. But "the guy who just came back to save the company is also selling shares" is not the confidence signal the bulls want it to be, and the market doesn't need much of an excuse to read it that way.

Here's the part that really stings, though: this is happening while the rest of the AI trade is on fire. Palantir ripped 16% in the past week. C3.ai, allegedly playing in the same enterprise-AI sandbox, mustered a limp 1.76%. When the tide is this high and your boat barely moves, that's not bad luck — that's the market telling you something about relative positioning, and right now it's telling you C3.ai isn't seen as a winner in this cycle, just a participant.

The bull case isn't fantasy — it's real, just diffuse. The AI infrastructure buildout is genuinely enormous; compute spend is already north of 1% of U.S. GDP and climbing. Enterprise AI adoption is a real secular trend, and C3.ai is nominally positioned inside it. Siebel's return could be read as institutional knowledge steering the ship through turbulence rather than a founder parachuting in to manage a slow leak. Consensus estimates even call for continued EPS improvement next quarter, to around -$0.26. Fine. But "narrower losses" is not a business model, and a company with 32%-plus short interest, a debt-to-equity ratio near 9, and revenue in freefall needs more than a founder's return and a friendly macro tailwind to earn a premium multiple.

Technically, the stock's hovering near $10.40, with support around $10.22 and a modeled path toward roughly $11.90 — an appealing chart if you squint, undermined by the fact that the forecasting model behind it has a directional hit rate barely better than a coin flip and a 51% error margin on the weekly view. Translation: don't trust the chart to bail out the fundamentals.

The next real answer comes September 9 — or September 2, depending on which calendar you believe, which is on-brand for a stock where even the earnings date is contested. That's when Q1 FY2027 numbers land, and it's the first real test of whether Siebel's second act is a genuine reset or a very expensive nostalgia tour.

Until then, my read is simple: this is a hold-your-nose situation, not a buy-the-dip one. The losses are shrinking, sure — but revenue is evaporating faster than the losses are improving, insiders are selling, and the analyst community can't agree on whether fair value is $6 or $16. When smart people are that far apart, the honest answer is nobody actually knows — and "nobody knows" is not where you want your conviction to live.