“AI depression” is seeping into psychotherapy clinics

“AI depression” is seeping into psychotherapy clinics

“Irreversible change” raises fears of its threat to the future of humanity

Americans have deep doubts about its developments

Talking about fears and anxieties related to artificial intelligence has become a regular part of Candice N.'s work day. Thompson, a psychotherapist who works in Silicon Valley, says her clients, who are often high-achieving tech employees, show practical concerns about the impact of artificial intelligence on their careers, as well as concerns about broader issues, such as its effects on the future of humanity, an issue that was widely discussed last month when a former expert in the field warned that the rapidly developing technology may destroy us all.

Artificial Intelligence...an existential threat

“In my career in therapy, I have never seen an existential threat like this,” says Thompson, who has been practicing psychotherapy for more than 17 years. “I don’t think any of us are immune from this risk, even my clients who don’t work in the technology field.” Other psychotherapists point out that concerns about artificial intelligence are becoming increasingly common among their clients, as this technology reshapes the work environment, consumes natural resources, and raises concerns about privacy and security.

Americans' deep "doubts."

A study conducted by the Pew Research Center - published last June - revealed that Americans view artificial intelligence with a great deal of “deep skepticism.” A larger percentage of adults expected that artificial intelligence would have a negative impact on them and society compared to a positive impact. In addition, about two-thirds of survey participants said they feel that AI technology is developing too quickly.

Artificial intelligence depression... after climate change depression

“People feel a lack of agency, a deep state of powerlessness, and a loss of control,” says Sherry Terkel, a sociologist, professor at the Massachusetts Institute of Technology (MIT) and author of Artificial Intimacy: Who We Become When We Talk to Machines.

She adds: “There is a similarity between this situation and depression caused by climate change. You feel that there is a deep imbalance, but you may see that any action will be futile, so depression turns into a general, vague condition.”

How does anxiety about artificial intelligence appear in psychotherapy sessions?

For people who already have diagnosed disorders — such as generalized anxiety disorder — the distress caused by AI may seem more severe. Emma Bastow, a 32-year-old mother of three young children who lives in Milton, New Hampshire, said she found herself thinking about artificial intelligence almost continuously for several days, after learning about the latest predictions portending the end of the world, and learning that some people are using artificial intelligence image generation tools to create non-consensual sexual content.

Basto, who suffers from obsessive-compulsive disorder, stated that she tends to slip into spirals of anxiety, because her mind deceives her that constant absorption in thinking will somehow guarantee her safety. Describing her condition, she said: “It's like having a thorn in your shoe; You may be busy doing other tasks, but it is still there, and it keeps poking you.”

She pointed out that her psychotherapist helped her break this vicious cycle by reminding her that there are things she simply has no control over, adding: “Worrying about these things will be of no use.”

Clinic visitors' fears of technology

Rachel Wood, a psychotherapist who studies the psychological effects of technology, regularly travels across the country to educate her fellow therapists about how people interact with artificial intelligence, explain the latest research on its effects, and discuss its presence in clinical practice.

She recently held a workshop with a group of psychologists in Oregon, and asked the audience whether reviewers raise existential concerns about artificial intelligence; About half of the therapists answered in the affirmative. Wood, who runs a small private practice in Colorado Springs, Colorado, said some of her clients raise this topic, too.

Sadness over “irreversible change”

She explained that their anxiety is often linked to feelings of sadness about “the irreversible change that we are currently experiencing, and the tremendous speed with which this change is occurring,” adding: “It is really difficult to keep up with this matter and comprehend its implications in light of this accelerating pace.” With artificial intelligence taking over tasks previously assigned to humans, many are asking: “What value do I provide?” Wood indicated that this shift in professional identity is one of the most prominent aspects that arouse in auditors a feeling of loss and uncertainty.

Between enthusiasm and the decline of critical thought

For her part, Yuexin Sun, a psychologist in Seattle, said that not all of her reviewers express concerns about artificial intelligence; Indeed, some of them, as she put it, are “very enthusiastic about artificial intelligence,” and view it primarily in a positive light. While others had more difficulty understanding what artificial intelligence means for them and their families.

For example, some software engineers who now ask AI to write code instead of doing it themselves have expressed to Liuxin that they feel they are missing out on the most interesting and important aspect of their work. Parents also expressed concerns that artificial intelligence would erode their children's critical thinking skills.

How do we stop this spiral?

There are no simple answers, but experts advise focusing on “the present moment, and doing what you can to make a tangible change, whether in your personal life or in public affairs.”

Basto uses the “5 Calls” app to reach out to elected officials to push for more regulatory safeguards for AI. As for psychotherapist Thompson, she calls for the ethical use of technology in the field of mental health. Last year, she presented a lecture during the “TEDx” activities entitled: “Why Can’t Artificial Intelligence Be Your Psychotherapist?” In which she highlighted the nature of psychotherapy based on human relationships, confidentiality, and governed by unique ethics and boundaries that guarantee the integrity of the therapeutic relationship.

“We are at a turning point where we can still exercise some control and make our voice heard on this issue,” Thompson says. While Sun points out that the feelings of existential anxiety expressed by some clients in psychiatric clinics often stem from their feeling of loss of control and lack of meaning. She adds that it can be helpful to ask, “What are the things I won't let AI do?” and she encourages her clients to view this stage as an opportunity to reevaluate what really matters to them.

Scientists have discovered a virus with a slightly human trait that appears to contain a gene that was stolen from our ancestors about 100,000 years ago. This virus likely uses our human genes to infect us.

Viruses hijack genes

Previous studies have demonstrated that viruses are able to pick up genes from their host organisms. However, the new discovery - which was published last week in the journal Science - may represent the first known case of a virus carrying a human gene that it acquired at a relatively recent stage in evolutionary history. “This affects us more directly,” says Nils Ilde, an evolutionary biologist at the University of Utah who was not involved in this research.

The virus's prowess at stealing DNA

The way viruses reproduce makes them particularly adept at stealing DNA: they invade a cell, then take control of its vital mechanisms by injecting their own genes into it. The infected cell is then forced to produce copies of viral genes, as well as proteins that coat these genes, to create new viruses. Eventually, the new viruses burst out of the infected cell, leading to cell death.

Sometimes this process may be marred by errors, and part of the host's DNA is integrated into the new viral genes. This molecular error may pose a threat to human health.

A virus... a bacterial gene

For example, viruses that infect bacteria can pick up genes that the bacteria used to resist antibiotics. Although these genes do not directly benefit the viruses themselves, they allow them to transfer this DNA to other bacteria that they later infect. This is one way antibiotic resistance spreads.

A virus... with a human gene

Now scientists have discovered a virus that acquired a gene from humans. This gene was not hidden, or hard to find; It was discovered by Cedric Feschotte, a biologist at Cornell University, and his colleagues while examining databases of viral genes. “I can't believe no one has made this discovery before us,” Vichot says. “Anyone can find these genes literally within a few minutes.”

Vichot and his colleagues discovered the human gene within the molluscum contagiosum virus, a virus closely related to the smallpox virus. This virus causes small blisters to appear on the skin, which usually disappear spontaneously after a few months. This virus only infects humans, and often infects children.

The human gene in question, called PC200, is a mysterious gene. It produces a ribonucleic acid (RNA) molecule in the brain, the function of which scientists have not yet precisely determined. Some researchers believe that the PC200 gene contributes to controlling the rate of protein production within brain cells.

Invasion and takeover 100 thousand years ago

The PC200 gene found in the molluscum contagiosum virus contains mutations that are not found in the version carried by humans. These mutations have accumulated since the virus picked up the gene from our ancestors, allowing researcher Vichot and his colleagues to estimate that the virus took over this human gene about 100,000 years ago.

The RNA molecule encoded by the PC200 gene has a unique structure that may sometimes cause our cells to insert it into a new piece of deoxyribonucleic acid (DNA). This piece is later integrated into our genome, creating an additional copy of the BC200 gene.

178 copies of the gene “PC200” in the human genome

It seems that this process was repeated many times. Vichot and his colleagues detected 178 copies of the PC200 gene in the human genome, which have accumulated over millions of years. It is possible that this defect is the reason for the gene being transmitted to the virus as well. Whatever its method of arrival, the BC200 gene has been present in the virus for the past 100,000 years. Vichot did not find any sample of the “molluscum contagiosum” virus that did not contain the “PC200” gene, which indicates that the virus takes advantage of this human gene to its advantage.

Although the original function of the PC200 gene may have been related to regulating proteins in our brains, the virus appears to have found a new function for it. Researcher Ilde suggested that the “molluscum contagiosum” virus may use our genes to control the process of producing proteins, a process known as “conversion and transmission.” “Poxviruses love to manipulate the transformation and transmission process,” Ilde added.

Are there similar viruses that carry human genes?

Virus-infected cells sometimes try to shut down their protein factories to stop producing new viruses, but poxviruses are able to use host proteins to restart those factories. This may pose a problem for humans; Our PC200 gene likely provides this service to the molluscum contagiosum virus when we become infected.

Finally, Vichot wondered whether he and his colleagues had accidentally found an extremely rare example of a human gene inside a virus, or whether this phenomenon was common in viruses. “Honestly, this makes me want to do more research,” he said.

Millions of years old, “remnants of viruses” may reveal the secret of leukemia patients’ response to chemotherapy. A new study has shown that patterns of activity of these ancient viral sequences within the genome are associated with poor response to treatment. Their inclusion in AI models has enhanced the ability to predict which patient is most likely to resist treatment.

The study was conducted by researchers from King's College London, who analyzed clinical and biological data for 271 patients with acute myeloid leukemia (AML), who received usual chemotherapy. The results were published in the journal Blood Neoplasia in August 2026, and indicated that what is known as the “dark genome” may contain additional indicators that will help in the future to improve treatment selection and develop more personalized strategies for leukemia patients.

When treatment doesn't work

Acute myelogenous leukemia begins in the bone marrow, can progress rapidly and interfere with the production of normal blood cells. Although a variety of tests are used to evaluate patients, doctors cannot yet accurately predict who will respond to chemotherapy and who will develop treatment-resistant disease.

In some cases, the patient does not achieve complete remission despite receiving treatment. The condition is known as treatment-resistant acute myelogenous leukemia, which makes choosing the appropriate treatment strategy a major challenge.

To investigate the possibility of improving prediction, the scientists developed three multi-layered machine learning models to which they gradually added different types of clinical and biological information. As for the final model, data related to the activity of ancient virus remnants found in the genome was added.

Viruses from the past inside our genes

These remnants are known as endogenous retroviruses (ERVs), and they are leftover parts of ancient viruses that infected human ancestors millions of years ago, then integrated into the genetic material and became part of the human genome.

These sequences are found mainly in non-coding parts of the genome, regions whose functions scientists still do not fully understand, and some of which are sometimes referred to as the “dark genome.”

These residues do not necessarily indicate the presence of an active viral infection, but their activity can affect a number of biological processes, including regulating the operation of genes and immune and inflammatory signals.

Artificial intelligence detects an additional signal

When the researchers added endogenous retroviral activity information to other clinical and molecular data, the model's ability to predict patients' response to chemotherapy improved.

The final model also recorded fewer false positives; That is, cases in which the model incorrectly predicts that a patient will develop treatment-resistant disease even though they are likely to respond to chemotherapy.

The researchers believe that these results indicate that information hidden in parts of the genome that often receive less attention may add value to traditional methods that focus on known genes.

Do the cells mimic a viral infection?

Scientists believe that one possible explanation is related to a phenomenon known as “viral mimicry,” where the cell behaves as if it is experiencing real viral activity. This may lead to the activation of immune and inflammatory signals within the cell; According to the hypothesis that the researchers are studying, these processes may help maintain some leukemia cells in a state that makes them more resistant to chemotherapy.

But the study does not yet prove that the activity of these viral sequences is the direct cause of treatment resistance, so this mechanism needs more research and experiments to confirm it.

Dr. Mohammad Mehdi Karimi, senior lecturer in bioinformatics at King's College London and lead author of the study, said that most cancer research focuses on traditional genes, while large parts of the “dark genome” remain less studied. He added that the results suggest the possibility of using a fingerprint derived from ancient viral sequences, along with traditional gene expression data, to help predict response to treatment.

A step towards more personalized treatment

The researchers stress that the results do not mean that the analysis of endogenous retroviruses has become a ready-made clinical test for making treatment decisions; Larger clinical studies are still needed to verify the model's accuracy and usefulness in different patient groups.

The researchers obtained patient data from the Beat AML trial, a landmark blood cancer precision medicine study launched in 2016.

The researchers believe that integrating “dark genome” signals with traditional clinical and genetic information may help in the future early identification of patients most susceptible to treatment resistance and support the development of more personalized treatment strategies for acute myelogenous leukemia.