Somewhere between 100 and 72 million years ago, while mosasaurs and plesiosaurs dominated the world’s oceans, something else was hunting in the deep: octopuses that may have stretched roughly 19 meters from arm tip to arm tip, longer than a bowling lane. That is the central claim of a study published in Science in June 2026, which describes two newly identified species of finned octopuses whose fossilized jaws were hidden inside rocks from Japan and Vancouver Island, Canada.
If the size estimates hold up, these creatures would rank among the largest invertebrate predators ever documented in the Cretaceous period, sharing the top of the food web with reptiles that have long been considered the era’s unchallenged rulers.
Beaks locked inside stone
The discovery hinges on an unlikely source of evidence: lower jaws, or beaks, preserved inside chunks of Cretaceous-age marine rock. Cephalopod beaks are made of chitin and protein, materials tough enough to survive fossilization even when the soft body around them decays completely. But finding them is another matter. These particular beaks were invisible to the naked eye, sealed within stone that looked featureless on the surface.
A team led by researchers in the Ikegami and Iba network at Hokkaido University extracted them using a process called grinding tomography. Each rock sample was shaved down layer by layer, photographed at every stage, and then reconstructed digitally. An AI model trained to recognize cephalopod jaw shapes flagged candidate structures within the 3D scans. The technique and its underlying workflow are documented in a separate Science report that focuses specifically on the imaging and AI pipeline the team calls “digital fossil-mining.” That companion paper details how the method works and how it was validated, while the primary Science paper uses the recovered fossils to describe the new species, estimate body size, and reconstruct ecological context. The pipeline had already proven effective in earlier work that recovered large numbers of cephalopod beaks from Cretaceous sediments and argued for a faster diversification of major cephalopod lineages than paleontologists had previously accepted.
This time, the pipeline turned up something far more dramatic.
Two new species, one startling size range
The recovered jaws were grouped into two species of finned octopuses in the order Cirrata: Nanaimoteuthis jeletzkyi and Nanaimoteuthis haggarti. Both are placed alongside modern dumbo octopuses of the genus Grimpoteuthis, the deep-sea animals known for the ear-like fins that give them their nickname, according to a Nature news analysis of the research.
To estimate body size, the researchers used beak-to-body scaling, a standard paleontological method that applies the proportional relationship between jaw dimensions and overall body length in living cephalopods to infer the size of extinct ones. The largest specimens pointed to animals reaching approximately 19 meters. For perspective, a modern giant Pacific octopus rarely exceeds five meters across, and today’s dumbo octopuses are typically smaller than a house cat.
Wear patterns on the fossilized beaks added another layer. The damage was consistent with shell-crushing behavior observed in modern cephalopods, suggesting these ancient octopuses were active predators capable of breaking hard-shelled prey rather than passive filter feeders or scavengers. According to the Nature news analysis, the Hokkaido University team described the discovery as evidence that “large, intelligent invertebrates occupied the top of some Cretaceous food chains,” sharing the seas with giant vertebrate hunters.
A deep-water range spanning two continents
The beaks come from marine sediments that formed in relatively deep offshore waters, consistent with the habitat of modern cirrate octopuses, which tend to live in cold, dark parts of the ocean. Their presence in both northwest Pacific deposits (Japan) and northeast Pacific deposits (Vancouver Island) indicates these animals were not local oddities. They ranged widely across ancient seas during a span of roughly 28 million years, from about 100 to 72 million years ago, a stretch that covers much of the Late Cretaceous.
That time window overlaps with the reign of mosasaurs, plesiosaurs, and large sharks. The co-occurrence of massive invertebrate and vertebrate predators in the same stratigraphic layers raises the possibility that they competed for resources, though the study stops short of claiming direct competition. Whether their diets overlapped, given that mosasaurs fed on fish, other reptiles, and ammonites, remains an open question.
Where the uncertainties lie
Independent experts have flagged several points that deserve caution, and the researchers themselves acknowledge limits to what beaks alone can prove.
The most consequential uncertainty involves the size estimates. Because no soft tissue survives, the 19-meter figure depends entirely on how closely the beak-to-body ratio in living octopuses applies to animals that lived tens of millions of years ago. If that ratio differed in ancient Cirrata, the estimates could shift significantly. The creatures might have been considerably smaller, or, less likely, even larger.
Classifying the beaks as cirrate octopuses also carries risk. No genomic data or preserved soft tissue exists to confirm the placement of N. jeletzkyi and N. haggarti within Cirrata. The assignment rests on morphological comparison of the jaws alone, and competing interpretations of beak shape could place the animals elsewhere in the cephalopod family tree. The Nature coverage notes that some paleontologists remain skeptical about how securely beak morphology can anchor such a classification.
The digital fossil-mining pipeline itself is still relatively new. Raw tomographic data and full details of the AI identification model have not been widely tested outside the research group’s own projects. Earlier applications lend credibility, but the specific accuracy rate and false-positive frequency of the AI component are not fully detailed in public summaries. Until other laboratories replicate the work on independent material, some caution is warranted.
There is also a conspicuous gap in the story: what happened to these animals. The youngest beaks date to roughly 72 million years ago, about six million years before the asteroid impact that ended the Cretaceous. Whether the giant octopuses were already declining before the mass extinction or were wiped out by it is unknown. The fossil record for soft-bodied cephalopods is notoriously sparse, and that gap may take years of additional grinding tomography to fill.
How to weigh the two Science papers and secondary sources
The strongest evidence is structural: the beaks exist, they were extracted from dated rock layers using a documented and previously validated method, and their morphology can be compared to living species. The primary Science paper (DOI: aea6285) provides the fossil descriptions, stratigraphic context, and scaling methodology that underpin every quantitative claim. The companion Science report (DOI: adu6248) documents the imaging and AI pipeline in detail. In short, the first paper describes the new species and builds the ecological argument; the second explains and validates the method that made the discoveries possible.
Secondary sources, including the Nature news piece, offer expert commentary that helps gauge where specialists agree or disagree with the authors’ interpretations. Institutional press releases, by contrast, tend to emphasize significance and novelty, sometimes using more dramatic language than the scientific text supports. Phrases like “apex predators” and “first direct evidence” are better understood as interpretive framing than as settled fact.
For now, the most defensible reading is this: large, shell-crushing cephalopods inhabited Late Cretaceous seas, and at least some of them were far bigger than any octopus alive today. The exact upper limit of their size, their precise branch on the octopus family tree, and their ecological relationships with giant marine reptiles remain active questions. As other research groups apply similar imaging methods to additional rock formations around the world, the picture should sharpen.
Until then, readers can reasonably accept the existence of these beaks and their Late Cretaceous age as solid facts while treating the 19-meter body length and detailed ecological reconstructions as informed but revisable estimates. In paleontology, especially when dealing with soft-bodied animals that fossilize poorly, provisional conclusions are the norm. The story of the Cretaceous giants is still being written, one painstakingly sliced rock at a time.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
