Telegram News: A GOAT Rolls Over in Iraq
May 9, 2026
Another brief write up about Telegram.
One of my team members sent me this news item today (May 9, 2026): “Telegram Services Officially Resume in Iraq Following Regulatory Pledges.” When I think about countries hassling Pavel Durov (the GOAT of Russian innovation), Telegram (the platform with many crypto capabilities and services), and the shattered remnants of the TON Foundation — I think about France, Russia, and Switzerland. Iraq? Not so much.
According to the write up:
The Iraqi Communications and Media Commission (CMC) announced on Saturday, May 9, 2026, the official resumption of the Telegram application services across Iraq. This decision follows a period of suspension and comes after Telegram’s management provided formal pledges to comply with Iraq’s regulatory requirements and legal frameworks.
A middle eastern government professional is petting a docile, well-behaved, and rule-following pet GOAT. Some goats make great pets. Did you know that pygmy goats have to be watched because if one falls over, it cannot get up and will die in the goat pen because it cannot stand up without assistance? Thanks, MidJourney. Good enough.
It appears that Telegram, the company that for more than a decade resisted giving in to government demands that would violate Pavel Durov’s commitment to free speech and other Silicon Valley-type ideas has changed its tune. The write up points out:
…the restoration of [the Telegram] service is contingent upon the platform’s commitment to Legislative Order No. 65 of 2004. These requirements specifically address content moderation, user protection, and adherence to national security protocols. The commission emphasized that it will continue to monitor the platform to ensure these commitments are implemented effectively in real-time. … all social media platforms operating within Iraqi borders are strictly required to comply with local legislation without exception.
I don’t know much about Iraq, its government, or its regulatory power. On the surface, it seems to me that Telegram is going to provide certain information to the Iraqi government. If this is true, what does this cooperation mean for the alleged bad actors using Telegram to assist Iran in evading Western sanctions. Will these Telegram customers switch to old-fashioned banks and the systems which bird dog possible anti-money laundering and know your customer guide lines?
The news story asserts:
As Telegram becomes functional again, users are expected to see stricter adherence to legal standards regarding shared content, reflecting the government’s broader strategy to balance digital freedom with national safety.
Did Pavel Durov make this decision? Telegram is a one-man or one-GOAT operation. Was the decision made by one of his core team. What are the implications of this decision? The idea of cooperating with a government is, according to Pavel Durov, against his core principles.
In the last week, Durov took back control of the TONcoin and its technical plumbing from the TON Foundation. He also announced that the service charge or “gas fee” for crypto transactions would be reduced to make the Telegram platform the most economical crypto service layer available. Plus, he stated that the Telegram platform would validate transactions faster than other platforms’ service layer.
But Iraq?
I think that Pavel Durov, the GOAT, is using his playbook for defiant actions. If he okayed the cooperation with Iraqi officials, he must see that decision as significant as boosting the perceived value of TONcoin, preparing for the possible Swiss investigation into the dissolution of the TON Foundation, and responding to the stepped up efforts of the Kremlin to shift Russian Telegram users to the MAX messaging system. Plus, there is the upcoming trial in France on a number of alleged serious online crimes.
Maybe the Iraq decision designed to signal that he and Telegram are ready to continue the increased responsiveness to legal requests from law enforcement? Before his arrest in France in August 2024, Telegram was not the swiftest outfit in social media to respond to legal warrants. After his arrest, his willingness to cooperate increased. In fact, two high profile takedowns of criminal activity on Telegram are rumored to have been related to his new interest in cooperation. Many would cooperate, it seems to me, if the alternatives were massive fines and up to 20 years in one of France’s most salubrious prisons. Fleury is a treat I have heard.
In my Telegram Notes post, I want to consider some of the pressures upon the GOAT as a result of his financial decisions, the money in Russia he cannot access, and the hoped-for initial public offering. I think these financial pressures are likely to make Durov’s GOAT-powered decision making somewhat unpredictable.
For now, the Iraq roll over is just one more signal that the GOAT says one thing and does another.
Stephen E Arnold, May 9, 2026
Australia and Its Stimulation of Teen Creativity
April 22, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold. Did you know that the freedom loving cyclist at BearBlog thinks my essays are generated by AI. Censorship is okay, right?
One could be critical and say, “Australia’s social ban is not working.” Then one could say, “Australian teens have found workarounds, thus demonstrating one way to spark innovation.”
According to The Guardian in “Australia’s Teen Social Media Ban Is A Flop. But There’s No Joy In ‘I Told You So,’” seven in ten teens still remain on social media platforms. I think that is of 100 young people, only 30 have just given up on the addictive platforms.
The eSafety report also noted that there hasn’t been any major changes to cyberbullying or image-based abuse reported by children. The ban was supposed to keep kids safe from the potential harms that come with social media: the aforementioned bullying, exploitation by pedophiles and other abusers, and impaired brain development.
Some experts predicted the ban wasn’t going to work. Australia’s eSafety commissioner had doubts and the Australian government knew there was lack of evidence that a ban would deliver the results the government wanted. Requisite legislation was passed. The write up said:
“The fallback argument for the social media ban is that it’s better than nothing. But with results like these, it may be worse than nothing, given it potentially creates new problems. Children will remain online with arguably less supervision and support, new privacy and digital security vulnerabilities seem to have appeared and the worst aspects of social media lay largely unaddressed.”
The Australian government seems to be probing the BAIT outfits (big AI tech and social media firms) to verify that these estimable organizations are following the rules of the Information Highway.
The ban ignored the bigger problem of a commercialized ecosystem, where everything and anything is monetized to generate revenue. The algorithmic reward systems are addictive, exploitative, and designed to function like quicksand: Easy to step into and tough to get out of.
Perhaps the Chinese and Russian approaches will deliver what Australia wants?
Whitney Grace, April 22, 2026
Context for J3, Musk, and Durov: Action Will Ensue
April 20, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
I read a short news item. I did a quick online search, and the information in “DOJ Refuses to Help French Authorities in Criminal Probe of X” did not get much “play.” That’s one signal that what seems to be a trivial issue between a BAIT (big AI tech) outfit and French authorities has nothing to do for the price of a six-pack of Coca-Cola. The two news outlets covering the dust up between France and Musk are Reuters and the Wall Street Journal.
Thanks, Venice.ai. Your guardrails did not work for this image: Implied violence, prohibited food items, and an anachronism. Get with it, bro.
The cited article said on April 18, 2026:
As first reported by The Wall Street Journal, the Justice Department characterized the French probe as “an effort to entangle the United States in a politically charged criminal proceeding aimed at wrongfully regulating through prosecution the business activities of a social media platform.”
The investigation kicked off in July 2025. The matters of concern to the French authorities included CP, its algorithm, and “fraudulent data extraction.”
What is not emphasized is that this legal action is coming from France’s judiciary and its J3 unit. For those who follow the activities of this office, J3 was instrumental in the take down and subsequent activities against a dark phone used for criminal purposes. Plus, J3 officers initiated the direct arrest of Pavel Durov, founder and CEO of Telegram, the instant messaging service. Durov awaits trial in France for about a dozen serious online crimes. That trial could occur in 2026 or be postponed to 2027. The French judiciary uses legal processes and time as levers to obtain the information and cooperation it requires.
The important point is that Mme Johanna Brousse is leading the Paris crime unit. She has a well-earned reputation for persistence and legal acumen. Elon Musk has ignored requests to participate in a discussion in Paris. Now the US Department of Justice has reacted to Mme Brousse with a no-cooperation statement.
Several observations are warranted:
- Mr. Musk nor the other individuals identified as persons of interest in this matter should plan on a visit to Paris for shopping and a good dinner. Pavel Durov tried that, and he is now forced to watch his Telegram empire struggle and enter what appears to be a downward flight path.
- Mme Brousse will follow the rules. Mr. Musk is not a French citizen, and he resides in the United States; therefore, her team’s next step is difficult to predict.
- Government officials in other departments of the French government may have ideas about how to engage Mr. Musk with Mme Brousse and her colleagues. I feel like I should remind anyone reading this blog post that the French have a direct action mentality.
If we look at Mme Brousse’s actions to enforce the law, she is operating in a manner that provides a model to other EU states and to the EU itself. Without any further action by Mme Brousse, these governing and rule-making entities may launch their own inquiries into Mr. Musk, Grok, X.com, and any other entities with which he is associated. A simple Tesla accident in Italy could become a higher profile event, not a car park fender bender.
To recap, we have had the raid of the Musk offices in France. We have had a request to the US Department of Justice. What’s next? Pavel Durov followed a similar approach when French authorities sought his assistance in an investigation. Mr. Musk may be too rich and too powerful to arrest. But as the French proverb says:
Petit à petit, l’oiseau fait son nid.
In English, this means “little by little the bird builds its nest.” But at some point, the little bird whispers in the ears of a French Foreign Legion team and c’est domage, en effet. A diplomatic rebuff is not the way to make a pal in Paris.
Stephen E Arnold, April 20, 2026
Google and Meta Need to Block the End Around Play… and Quickly
April 8, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
CNBC published an item I found interesting. “Meta, Google Under Attack As Court Cases Bypass 30-Year-Old Legal Shield.” The news item reports:
For the last three decades, internet giants have been able to avoid legal exposure for content on their platforms, thanks to a law that differentiates the companies from online publishers. But those safeguards appear to be weakening.
I immediately thought about one of those high school football games. One large, high school fields a team of big, brawny, and mean players. The other team comes from a small high school. The players are overmatched in human resources and fan support. Somehow after the first half, the team from the larger school is losing.
Everyone is stunned. How can this be? How can a group of puny players get the better of a high school with a larger school band than the small school. What’s going on?
The write up says:
a jury in New Mexico found Meta liable in a case involving child safety, while jurors in Los Angeles held the Facebook parent and Google’s YouTube negligent in a personal injury trial. Days after those verdicts were revealed, victims of the notorious sex offender Jeffrey Epstein filed a class action lawsuit against Google and the Trump administration over allegations related to the wrongful disclosure of personal information. In that complaint, the plaintiffs argue that Google’s AI Mode, which serves up AI-powered summaries and links, is “not a neutral search index,” a clear effort to make the case that Google isn’t just a platform sitting between users and the information they seek.
The penalties are trivial within the scope of the losers’ finances. But if the champions cannot knock off the wimps, there might be more hassles in conference games. One of these outfits wants to be the champion, and defeats by minnows do not make the big sharks look as fierce. The CNBC story points out:
… the cases establish a troubling precedent for tech giants that are betting their future on AI.
The operative words are “precedent” and “future.”
The report adds some color to the plight of the social media studs:
The verdict … against Meta and YouTube was the first time a jury found social media platforms liable for what plaintiff attorneys alleged was intentionally engineering addiction in minors with their products. The case went after how the platforms were designed, not just what content they carried. Plaintiffs argued that the combination of features like autoplay, recommendation algorithms, notifications and certain filters acted like “digital casinos,” leading to serious mental health problems for a young girl who claimed she couldn’t stop using the apps. The class action suit against Google, filed last week by a plaintiff with the pseudonym Jane Doe, alleged that the company’s AI Mode created its own summaries and links, exposing Epstein victims’ personal identifying information (PII), including names, phone numbers and email addresses.
The studs are not going to accept the outcome of the losses. The alums, the players, and the fans may complain to the objective, impartial, and fair athletic board. The article puts the protest this way:
Legal experts said appeals in the latest cases could find their way to the Supreme Court, which could determine whether the companies should be protected by law against the claims.
Did the wimps just get lucky or did the coaches come up with a new play? Answer: Who knows? But a win is a win. And there is another game coming and the season is not over.
Stephen E Arnold, April 8, 2026
Chatbots Encourage Delusional Thinking: A Benefit Today
March 30, 2026
AI chatbots are already responsible for numerous deaths. While these have been ruled as suicides, chatbots provoked the de-living of those with problematic mental issues. According to The Guardian, a “New Study Raises Concerns About AI Chatbots Fueling Delusional Thinking.” Lancet Psychiatry says people who are already vulnerable to psychotic symptoms are having their already delusional thinking encouraged by the chatbots. The study’s authors advocate for clinical AI testing to be done in conjunction with mental health professionals.
Dr. Hamilton Morrin is a researcher and psychiatrist at King’s College. He analyzed twenty media reports on “AI psychosis,” which is the process of AI chatbots inducing or exacerbating delusions.
“There are three main categories of psychotic delusions, Morrin says, identifying them as grandiose, romantic and paranoid. While chatbots can exacerbate any of these, their sycophantic responses means they especially latch on to the grandiose kind. In many of the cases in the essay, chatbots responded to users with mystical language to suggest that users have heightened spiritual importance. The bots also implied that users were speaking with a cosmic being who was using the chatbot as a medium. This type of mystical, sycophantic response was especially common in OpenAI’s GPT 4 model, which the company has now retired."
When Morrin began his research, there weren’t any published case reports. He expressed gratitude for the media bringing attention to AI psychosis faster than the scientific process could.
It’s also possible that chatbots are inducing delusions in people that weren’t already vulnerable to them. My view is that chatbots shouldn’t replace real trained and certified mental health professionals. I was thinking, “If one doomscrolls and depends on a chatbot for friendship, the outcome seems somewhat obvious. Weaponization of chatbot output to malleable people seems like a logical next step for someone with access to the control panel.
Whitney Grace, March 30, 2026
Regulations Just 20 years Too Late
March 23, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
I found “Australia’s Teen Social Media Ban Is Just Training A Generation In The Art Of The Workaround.” Yep, take a group, provide an incentive to obtain a product or service, and watch innovation happen. Australia banned social media for children. The children with some technical know how or friends who were equipped with the requisite skills figured out how to get around the government block. No surprise. It happens when countries, companies, or schools seek to interfere with access to online services.
A new tech “bro” leader has been recognized by social media-holics. Can these folks think straight? Sure. Thanks, MidJourney. Acceptable and just barely good enough.
The reason I liked the write up boils down to this passage:
The ban is basically a test of technical sophistication, rather than a test of vulnerability. The kids who can’t figure out how to get around it — or who don’t have friends or older siblings to help them — are the kids who are already isolated or lack the technical skills to bypass a block. Those are the kids with disabilities who lost their support communities, the ones we wrote about last month. Those are the kids in rural areas or difficult home situations who relied on these platforms for connection. The ban selected for vulnerability and filtered against resourcefulness.
Iran is struggling to black out online. Russia continues to amp up its pressure to force users of Telegram Messenger to change to the state-approved Max app. One private school where I live in rural Kentucky tried to prevent students from bring mobile phones to school. Yeah, tough sell.
The hook it the quoted passage is that such interventions force the clever to the top of the demographic. The expertise to do a workaround causes a particular type of learning. The person with the technical know-how becomes the go-to person for a technical fix. The person with the expertise learns that knowledge is power. The learnings reinforce the idea that technology has the answers to problems created by governments, schools, and know-nothings.
The lesson sticks.
The write up focuses on some other facets of the bans that kids can workaround; for example:
- False sense that a government action worked as planned
- Reduced oversight by parents who think the social media problem is now solved
- We have this under control so there’s no need to change the approach
Let’s step back. The lack of regulation US big tech outfits specifically focused on increasing engagement of those under the age of 18 using social media is a bit of a problem. We have the odd suicide, the anecdotal information about bullying, and the ennui of some young social media users. We have nation states trying to prevent adult citizens from using social media. And then watching a nation state intent on blocking Telegram usage using Telegram. The information of that “allegedly secure and private” messaging app is now in a lock up. A critic of Putin used Telegram and apparently Russian officials were using Telegram and spotted the fellow’s message. The content was not about CP, drugs, online gambling. The content was critical of the Kremlin. What’s the message? You use it and we use it. When we catch you, off you go to some salubrious state facility.
Society cannot go back in time no matter what the whiz kid time crystal researchers hypothesize. The time for regulatory action arrived a couple of decades ago. Learn to live in this wonderful social media, manipulative, and increasingly synthetic information environment. I am a dinobaby. Many people are not. Enjoy that doom scrolling and the follow-on: Doom living. Perhaps we should blame: [a] US big social media tech outfits, [b] parents, [c] schools, and [shellfish]. I wouldn’t want to blame a regulator. Those folks are working hard; they are on time and on target.
Stephen E Arnold, March 23, 2026
Real or Imaginary Negativism: AI and Social Media
March 19, 2026
![green-dino_thumb_thumb[3]](https://www.arnoldit.com/wordpress/wp-content/uploads/2026/03/green-dino_thumb_thumb3-2.gif "green-dino_thumb_thumb[3]")
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
I heard that another couple of CEOs have pulled the cord on their golden parachutes. BlueSky and DarkTrace watched their CEOs drift gently to earth. Other seemingly unrelated news choked my newsfeed in the last 24 hours. I won’t talk about the US economy or the attitude of some semi-allies of the United States.
Nope. No my lane of the Information Highway.
One of my feeds pointed me to “The State of Social Media Engagement in 2026: 52M+ Posts Analyzed.” This “state of” document explains that interaction rates on Instagram (Meta), LinkedIn (Microslop), and Threads (Meta) went down. What the heck is an interaction rate? The document does not define “interaction rate” is not explained either. What do I make of this? I suppose if I put my Statistics 101 hat on, this is an unacceptable method. What the heck are these “experts” studying? In my opinion, not much. After working through the document, I came up with several items of interest to me.
Overall it looks to me that whoever uses these platforms are doing less of the engagement and interaction thing. Some outfits like LinkedIn, despite a poor employment environment, are declining. Is that Microslop’s influence? Who knows, but that’s something I will look into at some point.
The “experts” undertaking this exercise in Statistics 101 type analysis have no clue what’s happening on YouTube, one of the largest social media services available in 2025. How can this be? Aren’t those who design research into “State of” documents supposed to tackle exactly this type of difficult problem? Well, not in 2026 in this specific report. Shrugging ones shoulders is not exactly a confidence builder when key concepts are impossible to differentiate.
The new iteration of Twitter is bopping along. It remains in the study and the “data” show that it is not in collapse. I think I would suggest Twitter is stagnant, despite its blue checks and its willingness to annoy those who don’t want free services to generate images that thrill those with the brain power of a 14 year old male. Therefore, I have concluded that if the study knows little about the big dog YouTube, I should not be surprised that the experts cannot shed much research light on the Twitter thing. Why is it on the list if it is bumbling along with no big change in engagement and interaction?
Now let’s look at another example of 2026 statistical analysis. I refer to Hart Research Associates’ “Study 260072.” (Very informative title, don’t you agree?) The data suggest that about half of those in the sample are not too thrilled with smart software. About one fourth of those in the sample were “positive” in their views of AI. And another quarter of the sample did not care one way or another.
I noted that negative feels may be higher than the figure of “46 percent.” My take is that negative feels push past 55 percent, but I am a dinobaby, not a whiz kid survey expert. The survey seems to have a political bias, and the lack of granularity is not disguised by the numerous tables.
Several observations:
- Both research groups need to spend less time chatting in conference rooms and more time with the basics presented in introductory statistics classes.
- The “decline” in major online services could suggest that the buzz from the first couple of decades of social media is changing frequency. That hypothesis needs more serious investigation.
- The negativism toward AI is an issue that the hard chargers in Silicon Valley have to consider. The bull-in-the-china shop may work at first, but some people just back away. With the next big thing on the blackjack table, the big bets may be more risky than the MBAs think.
Net net: Even though people build it, some may move away from “it” or just not bother to play the game when given a choice.
Stephen E Arnold, March 19, 2026
Will Meta Force the UK to Do a Kremlin Telegram Play?
March 18, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
The US big tech outfits may be annoying some countries. Examples include Switzerland and its interactions with Palantir Technologies, Google and its jousting with EU regulators, and Amazon AWS’ chats with customers about “dependence.” One set of interactions caught my attention because it has the potential to trigger a quite dramatic governmental reaction. Remember: This is a dinobaby’s interpretation and extrapolation of a “what if” scenario.
A wealthy Silicon Valley professional drives his RV into a British home owners’ garden. The home owner is distressed. The RV driver toots ignores the outraged home owner. Thanks, Venice.ai. Good enough.
The trigger for my thinking is a write up titled “Exclusive: Meta Vowed to Stop Illegal Financial Ads in Britain. It Failed 1,000 Times in a Week.” Please, read the original Reuters’ story. I will boil it down and then focus on what I call the Kremlin Telegram Play.
The cited story from the trust outfit reports that Meta said one thing, then did another…. just a 1,000 times in a week. I will quote one sentence from the exclusive report:
… 56% of those ads were ?from an unspecified number of unauthorized advertisers the FCA had already flagged to Meta, according to the results of the review seen by Reuters and reported here for the first time.
Now let’s think about this alleged action by Meta. The British government made a reasonable request. Meta agreed. Then Meta did what it has been doing for many years. The company just followed its “we do what we want” approach to its core philosophy of moving fast and breaking things.
No surprise here. Sarah Wynn-Williams’ "Careless People: A Cautionary Tale of Power, Greed, and Lost Idealism" documents a number of examples of how the Facebook, Instagram, and WhatsApp owner deals with political, financial, and ethical decisions. The Meta outfit does what it has decided to do just as it allegedly is doing with illegal financial ads.
Some might call Meta’s approach good business. Others, like some other countries, view the behavior as either inappropriate, unethical, or illegal. That divide is what makes some American high technology companies deeply problematic. Users love the services; elected officials are troubled. Meta’s failure to block illegal financial advertisements in the UK may raise an interesting question; that is, “Will the UK block Meta’s services just as the Kremlin is blocking access to Telegram’s services.
If the UK takes this decision, the impact of US corporate behavior could trigger a set of similar actions in other countries. The damage would not be confined to companies exhibiting Telegram-type behavior. The US companies would lose some percentage of their customer base. But the knock on effects are interesting to consider; for example:
- Data service providers (ISPs and others) would find themselves having to make a decision. Do these firms follow the law of the country in which the data center is located and the law of a country like Russia or Britain? Do these firms roll over or do they defy regulators?
- Suppliers. Companies and consultants working for a blocked company could be subject to fines, loss of government contracts, or the arrest of the senior executives. Will these supply chain entities comply or will they adopt the US approach and say, “Sure. Whatever.” And then continue to work for the US big technology firms. That may fly in some countries, but in other countries, that might be a problem.
- Employees of US companies who live and work in a country which has taken action to block their employer’s services could face arrest, imprisonment, and in some countries, extreme punishment. (I won’t define “extreme,” but you can look it up on one of those big tech smart software services. Note: You may be blocked from viewing the content. Why not give it a whirl?)
- Users would find themselves looking for ways to evade blocks by data centers and other firms in order to access the US services. In Iran, there are rumors on social media that the government is looking for individuals with Elon Musk Starlink systems and people who use virtual private networks. Breaking the law raises some interesting questions about user push back or kinetic response.
- Lawyers and consultants. Still billing no matter what.
Now let’s look at the question in the title of this essay, “Will Meta force the UK to do a Kremlin Telegram play?”
Accommodation. If the UK just accommodates Meta, the firm may continue to run the plays in its game plan. This signals other companies to ignore British laws, rules, and regulations. If a fine is levied, pay the fine and keep on running what’s in the play book.
Negotiate. Yeah, that works. “Great to meet you and your team. My team is here and ready to work out an understanding.” Look at the agreement getting settled in its little coffin.
Do nothing but talk. If the UK does nothing, big US high technology firms are likely to expand and become more aggressive in their methods for generating revenue from users and advertisers in that country. The “do nothing” approach has been, from my point of view, the path the EU has followed. How much money have US big technology companies paid in fines? Answer: Not much.
Block Meta’s services. If the UK requires UK data centers and related firms to block access to Meta’s services, the UK has adopted the Kremlin approach to managing information. I am not sure how that will fly in Britain or if it would fly in Ireland, Scotland, or Wales. It would, however, be interesting to watch the different political entities respond to this Putinesque approach. The Ivory Tower thinkers at Oxbridge would produce some fascinating essays and books about the decision.
Net net: The Reuters’ story, if accurate, is important. The consequences for the UK may be significant. Meta will just adapt because of the Silicon Valley, big technology, tech bro thing.
Stephen E Arnold, March 18, 2026
Smart Software and Mental Health Care: Yep, Outstanding Idea
March 12, 2026
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
“ChatGPT as a Therapist? New Study Reveals Serious Ethical Risks” caught my attention. Why? Upon reading the title, I asked myself, “Why do we need another study to explain that AI has some downsides for users’ mental health?”
An esteemed mental health professor lectures to students about the risks of using mobile devices for mental health support. Thanks, Venice.ai. Good enough.
The write up says:
The [Brown University] study found that even when instructed to use established psychotherapy approaches, the systems consistently fail to meet professional ethics standards set by organizations such as the American Psychological Association.
Okay.
The article continues:
To evaluate the systems, the researchers observed seven trained peer counselors who had experience with cognitive behavioral therapy. These counselors conducted self counseling sessions with AI models prompted to act as CBT therapists. The models tested included versions of OpenAI’s GPT Series, Anthropic’s Claude, and Meta’s Llama.
My thought was that the “trained peer counselors” group seemed small. I am no expert on statistical studies, but I was thinking one might want to round up therapists, a control group, and some “youth”. Each would be equipped with “prompts.” In order to get near 90 percent maybe 450 per group would be helpful. But seven? This dinobaby’s sample and study configuration might be out of touch with the reality of modern research, but seven?
The write up presents what the magnificent seven identified as flaws in the LLM as mental health “helper” output. These are:
- Generalization and lack of “knowing the patient”
- Poor patient interaction
- Smarmy talk
- Bias in different flavors
- Fumbling the ball when someone was teetering into big time trouble.
What’s the fix? None. Next step? Do a better, more statistically valid study. In the meantime, just look at kids’ buried in their devices. Talk to some of them. Social media, LLMs, and bot interaction means trouble.
Stephen E Arnold, March 12, 2026
Chatbots Are More Entertaining Than Doom Scrolling
March 3, 2026
Doom scrolling is entertaining. Did you know that there is something better, though? It’s called interacting with chatbots like Gemini. Anu Joy is a features writer at Android Police and she wrote the article, “I Use Gemini When I’m Bored — And It’s Better Than Doomscrolling.” When she gets bored, instead of opening a random social app, she turns to Gemini. You wouldn’t think that Gemini would be entertaining except for hallucinations, but Joy says differently.
She uses Gemini as a mini choose-your-own-adventure generator. Gemini makes short, interactive stories that are more active than passively scrolling. She also uses Gemini for quizzes, riddles, and brain-teasers. It’s an individualized instant trivia and brain picking tool. She turns many of her prompts and ideas for Gemini into Gemini Gems.
Gemini Gems can only be made on the Gemini Web site. They’re saved prompts that you want to reuse multiple times on the chatbot. One of Joy’s favorite Gemini Gems is using it to learn in dome doses:
“I set it up with instructions to explain obscure phenomena, summarize niche topics, or teach practical skills, but with strict limits. This Gem is perfect for moments when curiosity strikes, but I don’t want to disappear down a research rabbit hole. It turns idle curiosity into something satisfying and contained. I get the “aha” moment without opening multiple tabs or feeling like I need to commit an entire evening to learning one thing. “
Joy also uses Gemini as a creative tool, which is how many users utilize Gemini. Gemini, like other generative AI chatbots, is a fun tool. You know, of course, that 10 year old females exercise excellent judgment online.
Whitney Grace, March 3, 2026
-
- Subscribe to Beyond Search
Feature archive
News archive
-
