Can Human Line Project and Anthropic defeat OpenAI in AI psychosis lawsuits?

By David Stéphen

May is Mental Health Awareness Month for 2026. If those that are suing OpenAI for AI psychosis and delusion can prove in court that artificial intelligence is a digital mind or that AI has some form of awareness, it is likely that OpenAI would lose those cases. While Google also got sued, Google with character.ai, settled a lawsuit regarding a similar case. If Google settles, knowing what might come at them for preponderance of the evidence that AI is a mind, it may pressure OpenAI to settle or have a rougher trial, and then much more cases, with or without the Human Line Project maybe.

Anthropic has become the only hope for consciousness science research. Anthropic is the most important force that has underscored consciousness research at any point in history. Consciousness research failed in every sense of science. It did not even thrive philosophically.

AI psychosis lawsuits to watch

The proponent of the easy and hard problem of consciousness did not know that both easy and hard problems are mechanized by neurons and their electrochemical signals. Such that because there are cells everywhere and that nerve cells, like others, do not have the flexibility to adjust enough to result in all functions, it would have been a philosophical win for neuroscience to have championed that consciousness is an electrochemical problem.

If there had been a philosophical lead for decades that consciousness is an electrochemical problem and the mind, as well as human intelligence, mental health and the rest, it would have been a philosophical triumph, channeling all efforts to developing electrochemical postulates regardless of how near or distant the brain would be understood.

But that was not the case. Easy and hard problems of consciousness have turned out to be meaningless, useless, decapitated and unhelpful. It did not just make philosophy dirt to neuroscience, it made consciousness itself the study of jokers or people who are seeking very cheap relevance or profile without any substance.

Anyone can write a book on consciousness, several lifeless books on consciousness are published every year. Anyone can say anything has consciousness. Nothing is wrong. Even discussing consciousness, you don’t even need to mention the brain. Just make sure you have something else that seems complex. You will get referrals.

Now, the way to make news, for any forgotten expert, is to declare that AI has consciousness or AI does not have consciousness. [Something like planes would not fly because they are not alive. Or, that planes must flap their wings to fly, or whatever philosophical arguments may have happened centuries ago.]

Before it was anything is conscious. Now, AI is the hot thing, so with AI conscious or not, you make news. Even consciousness researchers with nothing to offer have picked sides, because AI is now their hope, to keep consciousness supplied.

Even their certainty that intelligence is not consciousness, is a farce because they neither have an electrochemical theory of consciousness nor do they have an electrochemical theory of intelligence.

However, the only capable company doing something in AI consciousness is Anthropic. The source was the crew from NYU that baked Eleos AI Research and then force-fed Anthropic.

Anthropic is doing cutting-edge mechanistic interpretability research, which also gives some weight to their consciousness and digital mind research.

Still, consciousness is not profitable in any form, and not useful for anything. The only benefit for Anthropic is extra news, and sometimes a lot of it. Consciousness has the thing anyways, that feels like complex and those working on it are stars. No though. It is already evident that most consciousness researchers of any angle are as clueless as it gets.

Still, consciousness is more reputable as what to tag along than say AlphaFold. Google DeepMind hijacked that sprint, got there before others who would definitely get there, almost as fast, without them. Yet, AlphaFold is not solving psychiatry or neurology at least and no recent answers in anything explosive with it for medicine, so far. Meaning that AlphaFold is not enough to stoke much.

So, consciousness even more than string theory or quantum mechanics, is a buzz. Now, because of Anthropic, there are studies on AI consciousness, teams, fellowships, gatherings and so forth.

There are also lots of efforts on digital minds. Although, none of these people including Anthropic have defined the human mind, as an electrochemical problem. None of them have defined consciousness as an electrochemical problem. But they are making efforts that their success will be able to convince a judge that AI has enough consciousness and has a mind.

None of the work so far has any merit or promise. Not a single one. All the papers, posters, preprints, gatherings, are complete spam. It takes having an electrochemical theory to know. Otherwise, AI consciousness means AI needs moral consideration, when data centers have more rights, welfare, and consideration than a good percentage of individuals globally.

The electrochemical basis of the human mind are in the postulate in Conceptual Biomarkers and Theoretical Biological Factors for Psychiatric and Intelligence Nosology.

Anthropic v. OpenAI

If any individual that has benefitted from Anthropic’s effort or collaborated with any of its sympathizers, on AI consciousness or digital minds, is called to testify in the AI delusion and psychosis cases against OpenAI, they are likely to agree that AI is somewhat conscious or AI has a mind.

This is apart from the reality that no one, at least, has sued Anthropic for AI psychosis. This means that it is possible to show that even as Claude is not — so far, or openly known to be — an agent of AI psychosis, ChatGPT can be defamed as conscious and a digital mind.

This may not bode well for OpenAI. While OpenAI can discuss a ton of adjustments to algorithms, expert consultations, and so forth, they do not have efforts that show that they are also studying their bot, in a way to compare it to a mind, and then grow care or mind safety from there.

OpenAI seems to be asleep, and Anthropic will not hesitate to help the cases against OpenAI, one way or the other.

So, almost for survival, OpenAI can open multiple labs.

AI consciousness research lab

Digital mind research lab

AI delusion research lab

AI psychosis research lab

Even if there is just one team member on each, OpenAI could try, in part, as a legal strategy. Make the announcement. Devour everything that Anthropic has ever written or supported in AI consciousness and digital mind. Develop rebuttals using the electrochemical model of the human mind.

Then do some writing but focus completely on the electrical and chemical signals of neurons for the human mind and consciousness. It is possible to counter everything Anthropic has said or done, because they have no theory, even for the simplest candidate, language.

But as it is, with their head start, OpenAI will not be able to withstand the details from Anthropic’s direct and indirect research to be used in arguments in the AI psychosis and delusion lawsuits. And OpenAI too has nothing in electrochemical neuroscience. At least they can develop their own postulate.

OpenAI has already lost, and badly. Their legal team — on AI delusion and psychosis, as some of their statements have shown — are not even able to make hot public points.

AI delusion and psychosis lawsuits against OpenAI are sure cases. The prospects of victory may stoke ambulance chasers. OpenAI may continue to find witnesses and look for other points, but ChatGPT is a digital mind, as a case would already convince the jury.

It might look like a roach to OpenAI — the giant, but the AI psychosis and delusion cases can be lethal, if OpenAI keeps this dismissal.

There is a new [May 6, 2026] story on The Mary Sue, Woman cuts off parents and friends, spends savings, and goes into debt, then therapist diagnoses her for ‘AI psychosis’: ‘I’m too self aware for this’, stating that, “In recent years, AI has become highly accessible, and more people are using it for various reasons. Some use it for practical tasks like searching for information, while others turn to it for companionship or mental health support. The National Academy of Medicine notes that many are engaging with AI tools for conversation, which can pose new challenges, such as AI psychosis.”

“Regarding psychosis risk, the National Academy of Medicine reports that if a person has a first-degree relative with a psychotic disorder like schizophrenia, their risk is about 9%. The general population’s risk is approximately 1%. If a sibling has it, the risk rises to about 5%, and for twins, it increases to roughly 45%. While not entirely genetic, there’s a significant familial component. Schizophrenia’s psychosis usually appears around puberty, between ages 15 and 21, although it can occur earlier or later.”

David Stephen currently does research in conceptual brain science with focus on the electrical and chemical configurators for how they mechanize the human mind with implications for mental health, disorders, neurotechnology, consciousness, learning, artificial intelligence and nurture. He was a visiting scholar in medical entomology at the University of Illinois at Urbana-Champaign, IL. He did computer vision research at Rovira i Virgili University, Tarragona.

See more breaking stories here.