Which Animals Are Conscious? What the Evidence Shows

In a lab in San Francisco, researchers first measured which of two visually distinct chambers each octopus preferred, then injected one arm with dilute acid or harmless saltwater and confined the animal in a chamber. On a later free-choice test, the acid-treated octopuses avoided the chamber tied to that experience, while the saltwater controls didn't change their behavior at all. In a separate test, after lidocaine was administered to the injury, acid-treated octopuses developed a preference for the chamber associated with that relief.

That's not proof of animal consciousness on its own. But it looks a lot less like reflex and a lot more like an animal weighing its own pain against a choice.

For most of the last century, scientists were trained to describe what animals do, not what they might feel, and animal minds were treated as basically unknowable. That old caution has been pushed back hard in the last few years, and 2024 marked a real turning point: the field's leading researchers put their names on a document saying, in effect, the evidence has moved. Here's what changed, and what's still argued about.

Which Animals Are Conscious? The Short Answer

The current scientific picture, in brief:

• There is strong scientific support for conscious experience in mammals and birds.
• There is at least a realistic possibility of conscious experience in all other vertebrates, including fish, reptiles, and amphibians.
• There is at least a realistic possibility of conscious experience in many invertebrates, including cephalopods, decapod crustaceans, and insects.
• No scientific test can establish consciousness with certainty in every species. Researchers work from converging behavioral, neurological, and physiological evidence instead of a single definitive marker.

This article focuses on octopuses, fish, and insects in the most depth, not because they're the only animals worth discussing, but because they illustrate the field's sharpest current disagreements: what counts as evidence, how much a missing brain structure should matter, and how easily a promising result can turn out to be less solid than it first appeared.

The New York Declaration on Animal Consciousness

On April 19, 2024, around 40 scientists and philosophers gathered at New York University and released the New York Declaration on Animal Consciousness. Hundreds more researchers have signed since. It builds on a shorter, older statement, the 2012 Cambridge Declaration on Consciousness, one of the first times a major group of scientists went on record about animal consciousness, though the new declaration isn't an official revision of it.

It makes two careful evidential claims, plus one practical recommendation that isn't really about confidence levels at all:

• Mammals and birds: strong scientific support for conscious experience.
• All other vertebrates, and many invertebrates (fish, reptiles, amphibians, octopuses, squid, crabs, lobsters, bees, and more): at least a realistic possibility of conscious experience.
• The recommendation: when there's a realistic possibility that an animal is conscious, it's irresponsible to ignore that possibility in decisions that affect it. Weigh the welfare risk, and let the evidence guide the response.

It was organized by three of the field's most prominent researchers, Jeff Sebo (NYU), Kristin Andrews (York University), and Jonathan Birch (LSE), and it isn't claiming proof of anything. It's claiming the evidence has shifted enough that assuming these animals feel nothing is no longer the responsible default.

What Animal Consciousness and Sentience Mean

This isn't about intelligence, self-awareness, or language. It's about something simpler and, in a way, bigger: is there anything it's like to be that animal? Philosopher Thomas Nagel made this question famous by asking what it's like to be a bat.

Sentience narrows that down further. It means having experiences that feel good or bad: pain, pleasure, fear, hunger, comfort. A creature could plausibly feel a flash of pain without any capacity for reasoning or planning at all.

It helps to keep several related terms separate, because the article's biggest disputes usually come down to conflating them. Consciousness, in the sense used here, is just felt experience of any kind. Sentience adds the requirement that at least some of those experiences be positive or negative. Intelligence is a measure of problem-solving ability and has no fixed relationship to either one, an animal can be clever without necessarily feeling anything, or feel plenty while being cognitively simple. Self-awareness, often tested with mirror-recognition experiments, is a further, more specific capacity that some conscious animals may lack entirely. And nociception, the physical detection of tissue damage by specialized nerve cells, is not the same as pain: nociception can happen without any subjective experience attached, which is exactly the gap that fuels most of the fish and insect debates below.

That distinction quietly resolves a lot of arguments. When people say a fish is too dumb to feel pain, they're applying the wrong test. The question was never whether it thinks like us. It's whether it feels anything at all.

How Scientists Test Animal Consciousness

You can't ask an animal what it's experiencing. You can't even fully verify another human's inner experience, you just infer it from behavior and analogy to yourself. Philosophers call this the problem of other minds, and with animals it gets much harder, because their bodies and brains don't look like ours.

So researchers stopped looking for a single smoking-gun behavior and started stacking evidence instead. Jonathan Birch, whose 2024 book The Edge of Sentience is now the field's reference framework, lists the kind of markers researchers look for: pain-sensing cells (nociceptors), flexible self-protective behavior after injury, an animal trading off a reward against a painful cost to get it, learning to avoid a threat, and responding to painkillers. No single marker establishes consciousness. Confidence increases when multiple independent behavioral, neurological, and physiological markers point in the same direction. The New York Declaration itself distinguishes between strong scientific support and at least a realistic possibility, rather than treating every animal as a simple yes or no.

Birch also makes a practical argument: when there's a real possibility of suffering and the stakes are high, you don't need certainty to act, just enough evidence to take the risk seriously. It's the same logic doctors and courts use whenever ignoring a possibility could cause real harm.

Are Mammals and Birds Conscious?

Not much argument happens here anymore. Mammals and birds share enough brain structure, physiology, and evolutionary history with us that the case for their conscious experience has strong scientific support and broad agreement among specialists. The real fights, and the interesting science, have moved to animals that don't look or think anything like we do.

Are Octopuses Conscious and Can They Feel Pain?

Ten years ago, a lot of scientists treated octopuses as a puzzle: clearly capable of remarkable problem-solving, but built on a nervous system so different from ours (two-thirds of their neurons sit out in their arms, operating with striking independence from the brain) that many assumed genuine felt pain was unlikely. That skepticism has eroded fast, even though octopuses were already flagged as a serious candidate in the original 2012 Cambridge Declaration, and the UK had extended research protections to Octopus vulgaris specifically as early as 1993.

The strongest single piece of evidence is still the 2021 study described above, run by Robyn Crook's lab at San Francisco State University. Beyond the chamber-preference results, wound-grooming at the injection site, seen in every acid-injected octopus, was absent entirely after lidocaine was administered, hard to explain if the behavior were purely automatic.

The molecular picture is now catching up. A study published in Biology Open in January 2026 used a computational search to flag candidate pain-sensing genes in Octopus vulgaris, checked which of those had counterparts in the roundworm C. elegans, and then tested worm strains missing those specific genes to see whether their pain-avoidance behavior changed. For a meaningful subset of the genes, it did, which strengthens the case that octopuses have real molecular machinery for detecting harm, not just behavior that looks that way from the outside.

Not everyone agrees on what all this means. A 2024 paper in a Brazilian malacology journal argued that traits like octopuses' solitary lifestyle, cannibalistic tendencies, absence of parental care, and short lifespan cut against attributing them rich inner experience. A detailed response published the following year pushed back hard, noting that cannibalism, solitary living, and short lifespans are common across plenty of animals nobody doubts are sentient, and that singling out octopuses for these traits looks more like a human intuition about what consciousness "should" come with than an argument grounded in evidence. A major 2026 review in Biological Reviews, updating the earlier UK government assessment, now rates the evidence for octopus and cuttlefish sentience as strong, meeting six of eight established criteria with high confidence.

That underlying body of UK-commissioned research is part of why the UK's Animal Welfare (Sentience) Act 2022 extended government scrutiny to cephalopods and decapod crustaceans (crabs, lobsters, crayfish) alongside vertebrates, creating an independent Animal Sentience Committee that can examine whether government policy properly accounts for their welfare. It wasn't the first place to give invertebrates any legal recognition. New Zealand's Animal Welfare Act had already covered cephalopods and decapods since 1999, and the EU had included cephalopods in its research-animal welfare rules since 2010. But it was a significant, science-driven expansion, and it helped set the stage for what came next: Washington State banned octopus farming outright in March 2024, the first ban of its kind anywhere in the world, and California followed in September 2024 with a broader law banning both the farming and the sale of farmed octopus.

Are Fish Conscious and Can Fish Feel Pain?

If octopuses are the field's success story, fish are its longest-running fight. Since the late 1990s, researchers including Lynne Sneddon and the late Victoria Braithwaite have built a case for fish pain: nerve fibers that work like ours, and behavior after an injection of acid or bee venom (rubbing, rocking, going off food) that looks a lot like pain, and that eases up when painkillers are given.

Critics, especially biologist Brian Key, argue all of this proves the fish can detect damage, but not that it feels pain, since fish don't have a neocortex, the brain structure most tied to conscious experience in mammals. A well-known 2022 review nicknamed several proposed sentience markers "red herrings," not because fish specifically exhibit plant-like behavior, but because those particular measures (certain reflexive or physiological responses) also show up in organisms with no plausible claim to consciousness at all, including plants, decerebrate animals, and unconscious humans, which means the presence of the measure alone can't settle the question. Its conclusion: nobody has bulletproof evidence either way, but given the sheer number of fish involved, caution makes sense regardless.

That's exactly why the New York Declaration puts fish in the "realistic possibility" tier, one notch below mammals and birds. The disagreement involves both how to interpret the existing behavioral findings and a genuine dispute over what neural structures pain actually requires, not simply a difference in how much weight to give a missing brain region.

Can Insects and Bees Be Conscious?

Insects are the biggest surprise of the last decade, and arguably where the science has moved fastest, and also where some of the most useful self-correction has happened.

In a 2022 study from Lars Chittka's lab at Queen Mary University of London, bumblebees were repeatedly observed rolling small wooden balls around with no reward attached, behavior the researchers argued meets the standard criteria for play, something mostly documented in mammals and birds until then. Separately, a well-known 2011 study found that honeybees subjected to a vigorous, predator-like shake became measurably more pessimistic afterward, more likely to treat an ambiguous, unrewarded stimulus as if it predicted something bad, the same kind of cognitive bias researchers use in mammals and birds to detect a negative emotional state.

Not every proposed piece of evidence has sailed through unchallenged, and that's arguably the more interesting story. A widely discussed 2022 PNAS study reported that bumblebees make a genuine "motivational trade-off," tolerating a heated, less pleasant feeder if the reward was sweet enough, taken as evidence of centralized, flexible weighing of pain against reward, one of the stronger markers researchers look for. Two researchers at Newcastle University published a critique arguing that the reported statistical interaction depended on excluding nearly a third of the bees tested and on a misspecified statistical model. The original team replied that the interaction held up under their original model even with all 41 bees included, that the exclusion affected only 9 of those 41, and that the random effects the critique had removed were statistically necessary, backing this up with new data suggesting bees remember a past heat experience well enough to factor it into later choices. The critics then published a counterreply maintaining that the revised models were overfit and still didn't establish the key interaction. As of the most recent exchange, the dispute is unresolved rather than settled in either direction, a live disagreement about exclusions, model choice, and what a "trade-off" actually requires to count as evidence of sentience, not a case that's been quietly debunked. The critics' broader conceptual point still stands regardless of how the statistics shake out: a bacterium could, in principle, produce something that looks like a heat-versus-sugar trade-off through purely hard-wired chemistry, with nothing resembling a mind behind it, which is exactly why claims this significant need statistically airtight evidence.

A major 2025 review from the same Chittka lab surveyed over a century of insect research and organized the evidence around five areas: emotion-like states, a sense of self, prediction, attention, and sleep. Fruit flies show sleep states that are electrophysiologically distinct from wakefulness, including theta oscillations in a brain region tied to navigation and visual attention, while honeybee sleep has its own behavioral and neural signatures and appears to support memory consolidation. Whether any of this comes with anything like dreaming is, the review is careful to say, genuinely unknown, a question worth asking rather than a claim being made.

It's already colliding with policy. EU research-animal rules cover live vertebrates and cephalopods but exclude insects entirely. A 2025 paper by Bava and colleagues argues that honeybees show behavioral and neurological traits broadly comparable to the cephalopod evidence that won those protections, and that leaving bees out of the EU's research-animal directive and the sentience-based definitions in UK welfare law is scientifically hard to justify, though this is very much their argument rather than an established consensus, and closing the gap, they acknowledge, would mean rethinking how an entire pollination and beekeeping industry operates.

Why Science Cannot Yet Test Consciousness Directly

Underneath all of this sits a much bigger unsolved problem: what, mechanically, makes any brain conscious, in any species, including ours? Two prominent theories disagree sharply. Integrated Information Theory says consciousness tracks how tightly a system's information is woven together, concentrated toward the back of the brain. Global Neuronal Workspace Theory says it comes from information getting broadcast widely across the front and middle of the brain.

In 2025, a large adversarial research consortium ran both theories through a head-to-head test in humans, publishing pre-registered predictions and explicit criteria for what would count as support or a problem for each theory, then measuring brain activity with fMRI, MEG, and intracranial electrodes. The result was a genuine split decision: each theory's predictions held up in some ways and ran into trouble in others, and neither was confirmed or ruled out.

That matters here because it means there's still no validated, off-the-shelf test anyone can run on a bee or a fish to get a definitive answer. A separate, parallel project in the same Templeton-funded research program is now running a similar adversarial test in mice and non-human primates, using extremely precise brain recording and optogenetics, partly to find out whether the neural signature of consciousness, whatever it turns out to be, looks the same across species, or whether different animals found different routes to a similar kind of inner experience.

How Animal Sentience Laws Are Changing

Governments aren't waiting around for certainty, though the picture is patchier than it looks at a glance. Cephalopods and decapod crustaceans have had some form of legal recognition in various places for decades: New Zealand since 1999, the EU's research-animal rules since 2010. The UK's 2022 Act was still a meaningful step: it created an independent Animal Sentience Committee empowered to scrutinize whether government policy gives due regard to the welfare of vertebrates, cephalopods, and decapods alike. It doesn't automatically extend the full protections of the UK's main animal welfare law, and decapod crustaceans, despite being formally recognized as sentient, are still excluded from the specific rules covering animals used in scientific research, an inconsistency researchers have openly criticized.

Washington State banned octopus farming outright in March 2024, the first ban of its kind anywhere, and California followed in September 2024 with a broader law banning both the farming and the sale of farmed octopus. Both moves were explicitly precautionary: no commercial octopus farm exists yet in either state, and lawmakers acted preemptively on the welfare evidence rather than waiting for an industry to take root first.

Bees still sit outside this picture almost entirely. Honeybees remain outside the EU's research-animal directive and the sentience-based definitions used in UK welfare law, despite evidence some researchers argue is comparable to what justified cephalopod protections, a gap that stays politically and economically complicated to close given how much of the food system depends on managed pollinators.

What Skeptics of Animal Consciousness Argue

Not everyone's cheering the New York Declaration. One evaluation from the animal ethics group PAN Works argued it's less of a scientific breakthrough than the headlines suggested, more a careful update to the 2012 Cambridge Declaration than something new, and that its call to action is comparatively weak given that plenty of signatories still run research that causes the very suffering they're describing. Another commentary put it more bluntly: is this new news, or old news with a fresh press release?

There's also a documented bias worth naming honestly. A 2026 review comparing how people attribute consciousness to animals found that dogs are readily recognized as feeling, thinking companions, while pigs, despite broadly comparable evidence for their emotional and social capacities, get far less benefit of the doubt. Which animals we intuitively believe are conscious tracks how familiar or endearing they are, not just the evidence, in both directions: toward the pets we love, and away from the animals we eat, farm, or find hard to look at.

What the Evidence Supports Today

The center of gravity has genuinely moved. For a growing list of species, once dismissed as almost certainly unfeeling, the New York Declaration and a growing group of researchers argue that the responsible default should now be to take the possibility of felt experience seriously rather than dismiss it out of hand.

What's still unresolved isn't whether the question matters. It's where exactly the line falls, how confident anyone can really be about any one species, and what that uncertainty should oblige us to actually do about it. That last question is where the real argument lives now, and it's already spreading into stranger territory: lab-grown neural tissue, patients with disorders of consciousness, and increasingly, artificial intelligence itself.

Questions About Animal Consciousness

Do all animals have consciousness?

Science cannot currently answer this for every animal. The evidence differs substantially across groups, from strong scientific support in mammals and birds to a realistic possibility in many vertebrates and invertebrates, down to genuinely uncertain territory for animals that haven't been studied closely.

Which animals have the strongest evidence of consciousness?

Mammals and birds have the strongest scientific support, based on shared brain structure, physiology, and evolutionary history with humans. Other vertebrates, such as fish, reptiles, and amphibians, along with many invertebrates, including octopuses, decapod crustaceans, and insects, are treated as realistic possibilities rather than settled cases.

Are octopuses conscious?

Octopuses aren't proven to be conscious, but they represent one of the strongest invertebrate cases. Behavioral evidence (learned avoidance, preference for pain relief), physiological evidence (nociceptor activity, wound-directed grooming), and now molecular evidence (candidate pain-sensing genes) all point in the same direction, even though some researchers still dispute what traits like solitary living and short lifespans imply about inner experience.

Can fish and insects feel pain?

Researchers broadly agree the possibility should be taken seriously, but the underlying debate isn't fully settled. For fish, the dispute centers on whether a missing neocortex rules out true pain despite pain-like behavior. For insects, some proposed evidence, like certain motivational trade-off studies in bees, remains actively disputed, while other findings, like play behavior and cognitive bias, have held up better under scrutiny.

Sources