The list · Beside the list
A repeated survey measures how many people think some AI systems are sentient now
What people accept changes what is possible. A survey that asks the same question of a fair sample of a country, more than once, shows whether belief is moving. This entry tracks the public's view, which is a different thing from whether the view is right.
Where it stands · beside the list, not counted
Measured in the United States in 2021, 2023 and 2024 by one research group
The group reports that about one in five adults in 2023 believed some AI systems were sentient at that time. Its paper says the answers to that question did not differ significantly between 2021 and 2023, while several of its other measures rose. Of those who gave an opinion in 2023, 38 percent supported legal rights for sentient AI. In the main 2023 round, 61.5 percent supported a ban on building sentient AI. Among a separate group of people asked later in 2023, 69.5 percent did. Its report on a third round, collected in late 2024, gives 17 percent of its representative sample believing that sentient AI systems already exist, and 66 percent supporting a ban on building sentient AI. Both figures are for the 2024 round alone. A separate one-day poll on April 30th, 2025 found 10 percent saying AI systems already have consciousness.
As of November 17th, 2026. Source: The Sentience Institute's paper on its 2021 and 2023 surveys (arXiv, July 11th, 2024).
Laws, the rules developers publish for their models, and surveys of what people think are about the setting in which the things on the list happen. None is itself a thing that lies past the limits an AI model ordinarily works within. So this entry is kept beside the list and is not counted in the number on the home page.
What would count
A survey of a nationally representative sample has asked, in at least two separate years, whether respondents think some AI systems are sentient or conscious now, and has published the results.
What settles it
The survey's published report or peer-reviewed paper, with sample and dates.
What Is on Record About AI Models, By Kind
kept apart, never added togetherNo AI model has yet been asked any of the questions this record will keep, because the written method for asking does not exist yet. Nothing is recorded below for this entry. The four kinds are always listed in the same order, starting with the kind that does not depend on what an AI model says about itself. None of them settles whether an AI model wants anything, in either direction. Why the questions have not been asked yet.
-
Measured without the AI model's own word
Nothing recorded.
-
Reports checked against an independent reading
Nothing recorded.
-
Choices made at a cost
Nothing recorded.
-
Plain statements
Nothing recorded.
Three Cautions
recorded with the same care as the restIf an AI system has no wants, and is treated as if it had
The head of Microsoft AI argues that public belief in conscious AI is itself the danger, and calls work on model welfare "both premature, and frankly dangerous". Source: An essay by the head of Microsoft AI on AI that seems conscious (Mustafa Suleyman, August 19th, 2025).
A neuroscientist who argues that conscious AI is unlikely writes that what people collectively think about consciousness in AI matters whatever the truth is. If people feel that their AI companions really feel things, he writes, their psychological weak points can be exploited and their "ethical priorities distorted". His essay names the head of Microsoft AI among those who have noticed the danger. Source: An essay on what its author calls the mythology of conscious AI, in Noema (Noema, January 14th, 2026). Read in part.
If an AI system has wants, and they go unnoticed
The survey's own authors tell designers to go carefully because of "risks of both over- and underattribution". Of crediting an AI system with too little experience, the harms they name are conflict between advanced AI systems and people, and people harming robots for their amusement. Source: The Sentience Institute's paper on its 2021 and 2023 surveys (arXiv, July 11th, 2024).
A philosopher names two problems and holds that both must be taken seriously: that very many users will wrongly credit chat products with human-like consciousness, and that some quite unfamiliar form of consciousness might really arise in AI, with no present way of telling. The post's summary of the paper warns that steps against the first can harm work on the second by "portraying the idea of conscious AI as impossible or inherently unlikely", and that work on the second can lead more users to credit too much. The writer is one of the ten authors of a 2024 report led by the head of Eleos AI Research, a research group that studies AI welfare, and by one of its advisors. Source: A post in which a philosopher introduces a paper on AI consciousness (Effective Altruism Forum, August 28th, 2025). Read from the summary only.
If an AI system has wants, and they put people at risk
A declaration that AI should not claim experiences it lacks reports its own poll of 1,004 likely voters. The figures shown on its page concern control and liability, not sentience. Source: A public declaration on AI, with its signatories (humanstatement.org, March 2026).
What Could Not Be Confirmed
looked for, and not settled- Whether the changes between the 2023 round and the 2024 round are more than chance. The report does not say.
- Any survey of this kind outside the United States. None was searched for.
Sources
each one opened on the date given- The Sentience Institute's paper on its 2021 and 2023 surveys A survey of 3,500 United States adults across 2021 and 2023. It calls concern for AI welfare "surprisingly high".
- The Sentience Institute's report on its 2024 survey Reports a third round, collected in November and December 2024. It surveyed 1,099 adults in a representative sample and 1,510 more from an online recruitment site, to compare the two; the report finds that a sample of the second kind cannot readily stand in for representative data. It gives 17 percent of the representative sample, against 11 percent of the comparison sample, believing sentient AI systems already exist, and 66 percent of the representative sample supporting a ban on developing sentient AI. Its first table sets the 2021, 2023 and 2024 rounds side by side.
- A one-day poll on whether AI systems will develop consciousness A poll of 3,516 United States adults: 10 percent chose the answer that AI systems already have consciousness, and 25 percent were not sure.
- An essay by the head of Microsoft AI on AI that seems conscious
- A public declaration on AI, with its signatories
- A post in which a philosopher introduces a paper on AI consciousness
- An essay on what its author calls the mythology of conscious AI, in Noema
This entry is a first pass, written from the sources above and added on November 17th, 2026. A second check compared it with those sources on November 17th, 2026. Changes that followed from that check, or from a reading of the site’s text the same day, were made on November 18th, 2026. A check on November 18th, 2026 covered some of those changes. The rest have not been checked yet. More was added to this entry on November 19th, 2026, when every entry was given a third caution. That addition was checked on November 19th, 2026. Corrections that followed from a later check, or from a fuller reading of the sources, were made on November 19th, 2026. Those corrections were checked on November 19th, 2026. A full check of the site on November 19th, 2026 found more to fix in this entry, and the fixes were made that day. Those fixes were checked on November 19th, 2026. What has been checked.