Datadog's $4B Upwind Bid Was Never Made, Israeli Outlet Says

Datadog offered nearly $4 billion for Upwind Security. Or it did not. The Information reported on 22 October that the $90 billion observability giant had made the approach — per someone with direct knowledge of the talks, as relayed by Dealroom — and Israel's Ctech answered the same day that "the reported offer was not made." Meanwhile Globes, reading the same report, said Upwind simply turned the offer down. Three versions of one deal. Neither company will say a word about any of them.

Here is the part nobody disputes: Upwind, a four-year-old Israeli cloud-security startup, is finalising a $300 million funding round at roughly a $3.8 billion valuation, led by TCV and Bessemer Venture Partners. So even if Datadog's bid is real, it carries almost no premium over what Upwind is already worth on paper. Which raises the only question that matters — why on earth would anyone sell for that?

We checked the figures ourselves this morning. Four billion dollars against a reported $50 million-plus of annual recurring revenue is roughly eighty times revenue. Against Upwind's new $3.8 billion private valuation, the bid is a premium of about five percent — the kind of number that does not get deals signed; it gets phones hung up.

Photo: Tel Aviv at night — Upwind was founded in Israel and keeps its R&D centre in Tel Aviv (its headquarters are in San Francisco). File photo. Credit: TFCforever / Wikimedia Commons, CC BY 3.0.

One bid, three stories

The Information's account, as relayed in detail by Dealroom News, runs like this: a near-$4 billion offer, recently made, with Upwind declining to accept and instead closing a $300 million round at about $3.8 billion. The report added that Upwind had fielded takeover approaches from several bidders in recent months and accepted none, and that Datadog had tried the same trick last year at roughly $1 billion before talks fizzled. That last detail checks out independently — Ctech reported the $1 billion talks in July 2025, when Upwind was valued at around $900 million.

Then came the Israeli rebuttal. Ctech, the tech arm of business daily Calcalist, wrote plainly: "CTech has learned that the reported offer was not made." No hedging, no anonymous qualification. And Globes, Israel's other financial paper, went the third way: the offer happened, and Upwind said no. All three cannot be true. And with Upwind declining to comment and Datadog not responding to requests for comment, nobody can tell you which one is.

This is how cloud-security M&A journalism works now: single-source whispers from one side of a negotiation, amplified across continents in hours, with the denial arriving before the lawyers have even drafted the denial. Investors, notably, did not shrug. Datadog shares fell 5.1% on the report — the market's way of voting that paying $4 billion for a company worth $3.8 billion on paper is not the kind of ambition it wants funded.

The maths: an 80x bid that barely clears the ask

Strip away the intrigue and the numbers tell their own story. Upwind raised $250 million in January at a $1.5 billion valuation. Less than a year later it is raising at $3.8 billion — more than a doubling, on revenue that The Information puts above $50 million a year, up from roughly $20 million at the end of 2025. (One honest caveat: that ARR figure rests on a single source, and Upwind does not publish its numbers.)

That trajectory is why a $4 billion bid fails the smell test as an offer, whatever Ctech says about its existence. Acquisitions usually carry a real premium — the buyer pays for control, for the future the founders are giving up. A five percent premium over the number your own bankers just put on the company is not an offer. It is a suggestion, and not a flattering one.

It also puts Upwind in rare air. Dealroom's analysis notes the implied multiple is richer than the one in Google's $32 billion deal for Wiz, the Israeli cloud-security giant whose sale reset every valuation in the sector. Whether or not anyone writes the cheque, Upwind is now being priced like the next Wiz — and the founders, who sold their last company, Spot.io, to NetApp for $450 million in 2020, have shown they know exactly how this game is played.

Why Datadog keeps circling

Datadog is not buying randomly. The company has spent the last two years muscling from observability into security — Cloud SIEM, code security, cloud security posture management — and co-founder and CEO Olivier Pomel has been unusually candid about the strategy. "We've learned how to properly go to market for security," he said on Datadog's Q3 2025 earnings call. "The combination of our underlying platform strength, technologies like FlexLogs, and our maturing product portfolio is starting to click in a way that's very exciting."

At a conference in September, Pomel went further: companies, he said, "will need more integrated systems that can respond at 'machine speed,' rather than relying on numerous separate products feeding alerts to human reviewers," according to The Markets Daily. About a quarter of Datadog's customer base now uses its security products. Upwind — runtime-first, AI-agent-aware, built to watch the exact workloads Datadog already monitors — would slot into that integrated platform like it was designed for it. Which, in a sense, it was: these two companies have been dancing around each other for over a year.

The price has moved while Datadog hesitated. The $1 billion of July 2025 is the $4 billion of October 2026, and the lesson for acquirers is brutal: in this market, waiting a year costs you three billion dollars. Datadog reports third-quarter results on 4 November, and analysts will be listening for any hint of how far the security push goes. There is also delicious symmetry in the background — The Information reported in August that Palo Alto Networks' chief executive had explored acquiring Datadog itself, a reminder that Palo Alto Networks has had its own bruising run at the intersection of security M&A and market panic.

What Upwind actually is

Upwind was founded in 2022 by chief executive Amiram Shachar with fellow Spot.io alumni, including chief technology officer Tal Zur and Lavi Ferdman. The name is a nod to Shachar's passion for wing foiling, per CyberPro Magazine — a small human detail, but a telling one: these are founders who like speed and catching the wind early.

The product watches what actually happens inside live cloud environments — runtime monitoring rather than periodic scans of stored configurations — and increasingly watches AI agents too: the autonomous software that connects directly to cloud infrastructure, business systems and production data. Customers named in reporting include Roku and Brazil's Nubank. Upwind does not publish customer counts, so treat the roster as illustrative rather than complete.

In September, Upwind made its own acquisition: Aegis, an AI-security startup barely nine months old, bought in an all-stock deal reportedly worth tens of millions. Aegis founders Omri Limor and Saar Ankonina — both veterans of Israel's Unit 8200 intelligence corps — now lead the new Upwind AI Labs, researching attacks on AI agents, skills and plugins. "AI security is developing extremely quickly, and by acquiring Aegis and bringing Omri and Saar on board, we are building a team that can move just as quickly," Shachar said, in comments reported by The Next Web. The timing explains the valuation: enterprises are discovering that AI models are now finding critical flaws on their own, while attackers are turning AI agents into attack tools — and every buyer wants the company that guards the agents.

Photo: A server room at KTH Royal Institute of Technology's Parallel Computer Center, Stockholm — the cloud infrastructure Upwind's runtime platform watches over. File photo. Credit: Esquilo / Wikimedia Commons, CC BY-SA 3.0.

Our take: the denial is the least interesting part

Let's be blunt. A $4 billion bid that carries a five percent premium is either a colossally cheap takeover attempt or a shoddy piece of single-source deal gossip dressed up as an exclusive. The truth probably sits somewhere boring in the middle: Datadog talked to Upwind, a number near the current valuation was floated, Upwind's bankers declined to be flattered, and three newsrooms turned one phone call into three mutually exclusive stories. That is not an information failure — it is the deal market working exactly as it does now, at machine speed, with everyone leaking.

But the more important story survives whichever version is true. Private markets have priced Upwind so aggressively that strategic acquirers can barely construct a bid that clears the ask. Datadog waited fifteen months and the price quadrupled. Palo Alto Networks reportedly circled Datadog. Google paid $32 billion for Wiz. The Israeli cloud-security pipeline is now a seller's market so lopsided that the buyers look like they are bidding against the future — because they are. Whether this particular $4 billion existed or not, the next one will be bigger, and Upwind's founders have already shown they would rather raise than sell.

What happens next

The $300 million round is the near-term event. Once it closes, Upwind will have raised well over $700 million since 2022 and will be under no financial pressure to entertain anyone — Datadog included. The next scheduled data point is Datadog's third-quarter earnings on 4 November, when Pomel will either be asked about Upwind or will conspicuously not be. Watch that call: the shape of his non-answer will tell you more than any of this week's reporting.

And if The Information's sources were right after all, expect Datadog to come back. A company does not circle the same target twice, fifteen months apart, because it has changed its mind about the strategy. It comes back because the target keeps getting more expensive, and the regret keeps compounding.

Did Datadog really offer $4 billion for Upwind?

Unconfirmed. The Information reported the offer via sources with direct knowledge of the talks; Ctech says its own reporting shows no offer was made; Globes says the offer happened and Upwind rejected it. Neither company has commented publicly. Treat the bid as reported, not established.

Why would Upwind turn down $4 billion?

Because it is barely above the $3.8 billion valuation of the funding round it is closing right now — roughly a five percent premium, which in acquisition terms is pocket change. With revenue reportedly tripling this year and fresh capital incoming, the founders have every incentive to keep building. They have exited before; they are in no hurry.

What does "80x ARR" actually mean?

It means the buyer would pay eighty times Upwind's annual recurring revenue — about $4 billion against a reported $50 million-plus. Multiples that high are a bet on explosive future growth, not current sales. Dealroom notes it is richer than the multiple in Google's $32 billion Wiz deal. One caveat: the $50 million ARR figure comes from a single source.

What is Upwind AI Labs?

A new research unit Upwind created in September 2026 after acquiring AI-security startup Aegis in an all-stock deal reportedly worth tens of millions. Led by Aegis founders Omri Limor and Saar Ankonina, both Unit 8200 veterans, it studies emerging attacks on AI agents, skills and plugins — the threats that appear when autonomous software starts acting inside company infrastructure.

Sources: The Information (original exclusive, via Dealroom News); Ctech/Calcalist; Globes (via Dealroom and Sunburst Markets relays); TradingView/Seeking Alpha; Stocktwits; The Markets Daily; The Next Web via WebProNews; beinsure; CyberPro Magazine.

About the author: Daniel Rao covers cybersecurity, cloud infrastructure and the business of technology for TechPulse USA.