How Much Do Fake Account Verifications Really Cost Across the World?

A new site that tracks the daily fluctuating costs behind building a bot army on over 500 social media and commercial platforms – from TikTok to Amazon and Spotify – in every nation on the planet is launched by the University of Cambridge.

For the first time, the Cambridge Online Trust and Safety Index (COTSI) allows the global community to monitor real-time market data for the “online manipulation economy”: the SIM farms that mass-produce fake accounts for scammers and social bots.

Image: DIW-AIgen

These markets openly sell SMS message verifications for fake profiles across hundreds of sites, providing a service for “inauthentic activity” ranging from vanity metrics boosts and rage-bait accounts to coordinated influence campaigns.

A new analysis using twelve months of COTSI data, published in the journal Science , shows that verifying fake accounts for use in the US and UK is almost as cheap as in Russia, while Japan and Australia have high prices due to SIM costs and photo ID rules.

The average price of SMS verification for an online platform during the year-long study period running to July 2025 was $4.93 in Japan and $3.24 in Australia, yet just a fraction of that in the US ($0.26), UK ($0.10) and Russia ($0.08).*

The research also reveals that prices for fake accounts on Telegram and WhatsApp appear to spike in countries about to have national elections, suggesting a surge in demand due to “influence operations”.

The COTSI team, based in Cambridge’s Social Decision-Making Lab, includes experts in misinformation and cryptocurrency. They argue that SIM card regulation could help “disincentivise” online manipulation, and say their tool can be used to test policy interventions the world over.

The team suggest that platforms should add labels showing an account’s country of origin for transparency, as recently done on X, but also point out such measures can be circumvented – a service provided by many vendors in the study.

“We find a thriving underground market through which inauthentic content, artificial popularity, and political influence campaigns are readily and openly for sale,” said Dr Jon Roozenbeek, study co-lead and senior author from the University of Cambridge.

“Bots can be used to generate online attention for selling a product, a celebrity, a political candidate, or an idea. This can be done by simulating grassroots support online, or generating controversy to harvest clicks and game the algorithms.”

“All this activity requires fake accounts, and each one starts with a phone number and the SIM hardware to support it. That dependency creates a choke point we can target to gauge the hidden economics of online manipulation.”

Co-lead author Anton Dek, a researcher at the Cambridge Centre for Alternative Finance, said: “Misinformation is subject to disagreement across the political spectrum. Whatever the nature of inauthentic online activity, much of it is funnelled through this manipulation market, so we can simply follow the money.”

“A sophisticated bot can run an influence campaign through hundreds of fake accounts” - Dr Jon Roozenbeek.

Murky global market

To register a new account, online platforms require SMS (Short Message Service) verification: a text message containing a code sent to a valid phone number. This is intended to confirm a human is setting it up.

Over the last decade, a murky global marketplace has emerged with the infrastructure to bypass this security protocol, and automatically generate and sell fake accounts in bulk.

Companies claiming to offer privacy solutions operate “farms” of thousands of SIM cards and SIM banks – both real and virtual – to provide SMS verifications and re-route web traffic though mobile networks to disguise its origin.

Fake accounts bought from this “transnational grey market” of informal businesses, often based in jurisdictions with little legal oversight, are central to online scams.

This market is also behind many malicious bot campaigns now dominating propaganda and PR dark arts, according to Cambridge researchers. “A sophisticated bot can run an influence campaign through hundreds of fake accounts,” said Roozenbeek.

“Generative AI means that bots can now adapt messages to appear more human and even tailor them to relate to other accounts. Bot armies are getting more persuasive and harder to spot.”

For example, a study last year uncovered a botnet of 1,140 accounts on X using generative AI to run automated conversations.

Fake account index

The team built COTSI with opensource data pulled from some of the world’s biggest fake account suppliers. Researchers identified seventeen vendors and sorted by traffic to focus on the top ten. Four of these are used at any one time to construct the global price index, with others kept in reserve.

Importantly, COTSI monitors not just prices but also the available “stock” of fake accounts listed by each vendor in every country for hundreds of platforms.

These include all social media channels, as well as cash, dating and gaming apps, cryptocurrency exchanges and sharing economy sites such as AirBnB, music and video streaming services, ride-hailing apps such as Uber, and accounts for major brands such as Nike and McDonald’s.

“One SIM card can be used for hundreds of different platforms,” said Dek. “Vendors recoup SIM costs by selling high-demand verifications for apps like Facebook and Telegram, then profit from the long tail of other platforms.”

Additional analyses show global stocks of fake accounts are highest for platforms such as X, Uber, Discord, Amazon, Tinder and gaming platform Steam, while vendors keep millions of verifications available for the UK and US, along with Brazil and Canada.**

Meta, Grindr, and Shopify rank among platforms with the cheapest fake accounts for sale, at a global average of $0.08 per verification. This is followed by X and Instagram at an average of $0.10 per account, TikTok and LinkedIn at $0.11, and Amazon at $0.12.

The researchers tested the market themselves, with mixed results. Attempting to verify fake US Facebook accounts only worked 21% of the time with one big provider, but over 90% with another. Much of this difference comes down to virtual versus physical SIMs.***

“Fingerprinting by some platforms can mean IP addresses get banned if registration fails,” said Dek. “High-quality verifications involve a physical SIM, requiring huge banks of phones. Nations in which SIM cards are more expensive have higher prices for fake accounts. This is likely to suppress rates of malicious online activity.”

“The COTSI shines a light on the shadow economy of online manipulation by turning a hidden market into measurable data”, - Prof Sander van der Linden.

Pre-election prices

To investigate if political influence operations can be seen in these markets, the team analysed price and availability of SMS verifications for eight major social media platforms in the 30 days leading up to 61 national elections held around the world between summer 2024 and the following summer.****

They found that fake account prices shot up for direct messaging apps Telegram and WhatsApp during election run-ups the world over, likely driven by demand. An account on Telegram increased in price by an average of 12%, and by 15% on WhatsApp.

Accounts on these apps are tied to visible phone numbers, making it easy to see the country of origin. As such, those behind influence operations must register fake accounts locally, say researchers, increasing demand for SMS verifications in targeted nations.

However, on social media platforms like Facebook or Instagram, where no link between price and elections was found, fake accounts can be registered in one country and used in another. They also have greater reach which keeps demand high.

“A fake Facebook account registered in Russia can post about the US elections and most users will be none the wiser. This isn’t true of apps like Telegram and WhatsApp,” said Roozenbeek.

“Telegram is widely used for influence operations, particularly by state actors such as Russia, who invested heavily in information warfare on the channel.” WhatsApp and Telegram are among platforms with consistently expensive fake accounts, averaging $1.02 and $0.89 respectively.

‘Shadow economy’

The manipulation market’s big players have major customer bases in China and the Russian Federation, say the research team, who point out that Russian and Chinese payment systems are often used, and the grammar on many sites suggests Russian authorship. These vendors sell accounts registered in countries around the world.*****

“It is hard to see state-level political actors at work, as they often rely on closed-loop infrastructure. However, we suspect some of this is still outsourced to smaller players in the manipulation market,” said Dek.

Small vendors resell and broker existing accounts, or manually create and “farm” accounts. The larger players will provide a one-stop shop and offer bulk order services for follower numbers or fake accounts, and even have customer support.

A 2022 study co-authored by Dek showed that around ten Euros on average (just over ten US dollars) can buy some 90,000 fake views or 200 fake comments for a typical social media post.

“The COTSI shines a light on the shadow economy of online manipulation by turning a hidden market into measurable data,” added co-author of the newSciencepaper Prof Sander van der Linden.

“Understanding the cost of online manipulation is the first step to dismantling the business model behind misinformation.”

---

*The data used in the study published in Science, as well as the additional analyses, was collected between 25 July 2024 and 27 July 2025.

** In April 2025, the UK became the first country in Europe to pass legislation making SIM farms illegal. Researchers say that COTSI can be used to track the effects of this law once it is implemented.

*** Lead author Anton Dek explains: “By virtual SIM, we mean virtual phone numbers typically provided by Communications Platform as a Service (CPaaS) or Internet-of-Things connectivity providers.”

“These services make it easy to purchase thousands of numbers for business purposes. Such numbers are usually inexpensive per unit, but they often carry metadata indicating that they belong to a CPaaS provider, and many platforms have learned to block verifications coming from them. On the other hand, when a physical SIM card (or eSIM) from a conventional carrier is used, it is much harder to distinguish from a normal consumer’s number.”

**** The platforms used were Google/YouTube/Gmail; Facebook; Instagram; Twitter/X; WhatsApp; TikTok; LinkedIn; Telegram.

***** A recent law passed by the Russian Federation banned third-party account registrations, which saw vendors suspend SMS verification registered in Russia alone as of September 2025. However, this has not stopped vendors operating from Russia offering services linked to other nations.

---

This article was originally published by the University of Cambridge on December 11, 2025. Written by Fred Lewsey. Edited by DIW staff. Licensed under Creative Commons Attribution 4.0 International License.

by Web Desk via Digital Information World

AI’s errors may be impossible to eliminate – what that means for its use in health care

In the past decade, AI’s success has led to uncurbed enthusiasm and bold claims – even though users frequently experience errors that AI makes. An AI-powered digital assistant can misunderstand someone’s speech in embarrassing ways, a chatbot could hallucinate facts, or, as I experienced, an AI-based navigation tool might even guide drivers through a corn field – all without registering the errors.

People tolerate these mistakes because the technology makes certain tasks more efficient. Increasingly, however, proponents are advocating the use of AI – sometimes with limited human supervision – in fields where mistakes have high cost, such as health care. For example, a bill introduced in the U.S. House of Representatives in early 2025 would allow AI systems to prescribe medications autonomously. Health researchers as well as lawmakers since then have debated whether such prescribing would be feasible or advisable.

How exactly such prescribing would work if this or similar legislation passes remains to be seen. But it raises the stakes for how many errors AI developers can allow their tools to make and what the consequences would be if those tools led to negative outcomes – even patient deaths.

As a researcher studying complex systems, I investigate how different components of a system interact to produce unpredictable outcomes. Part of my work focuses on exploring the limits of science – and, more specifically, of AI.

Over the past 25 years I have worked on projects including traffic light coordination, improving bureaucracies and tax evasion detection. Even when these systems can be highly effective, they are never perfect.

For AI in particular, errors might be an inescapable consequence of how the systems work. My lab’s research suggests that particular properties of the data used to train AI models play a role. This is unlikely to change, regardless of how much time, effort and funding researchers direct at improving AI models.

Image: DIW-AIgen

Nobody – and nothing, not even AI – is perfect

As Alan Turing, considered the father of computer science, once said: “If a machine is expected to be infallible, it cannot also be intelligent.” This is because learning is an essential part of intelligence, and people usually learn from mistakes. I see this tug-of-war between intelligence and infallibility at play in my research.

In a study published in July 2025, my colleagues and I showed that perfectly organizing certain datasets into clear categories may be impossible. In other words, there may be a minimum amount of errors that a given dataset produces, simply because of the fact that elements of many categories overlap. For some datasets – the core underpinning of many AI systems – AI will not perform better than chance.

For example, a model trained on a dataset of millions of dogs that logs only their age, weight and height will probably distinguish Chihuahuas from Great Danes with perfect accuracy. But it may make mistakes in telling apart an Alaskan malamute and a Doberman pinscher, since different individuals of different species might fall within the same age, weight and height ranges.

This categorizing is called classifiability, and my students and I started studying it in 2021. Using data from more than half a million students who attended the Universidad Nacional Autónoma de México between 2008 and 2020, we wanted to solve a seemingly simple problem. Could we use an AI algorithm to predict which students would finish their university degrees on time – that is, within three, four or five years of starting their studies, depending on the major?

We tested several popular algorithms that are used for classification in AI and also developed our own. No algorithm was perfect; the best ones − even one we developed specifically for this task − achieved an accuracy rate of about 80%, meaning that at least 1 in 5 students were misclassified. We realized that many students were identical in terms of grades, age, gender, socioeconomic status and other features – yet some would finish on time, and some would not. Under these circumstances, no algorithm would be able to make perfect predictions.

You might think that more data would improve predictability, but this usually comes with diminishing returns. This means that, for example, for each increase in accuracy of 1%, you might need 100 times the data. Thus, we would never have enough students to significantly improve our model’s performance.

Additionally, many unpredictable turns in lives of students and their families – unemployment, death, pregnancy – might occur after their first year at university, likely affecting whether they finish on time. So even with an infinite number of students, our predictions would still give errors.

The limits of prediction

To put it more generally, what limits prediction is complexity. The word complexity comes from the Latin plexus, which means intertwined. The components that make up a complex system are intertwined, and it’s the interactions between them that determine what happens to them and how they behave.

Thus, studying elements of the system in isolation would probably yield misleading insights about them – as well as about the system as a whole.

Take, for example, a car traveling in a city. Knowing the speed at which it drives, it’s theoretically possible to predict where it will end up at a particular time. But in real traffic, its speed will depend on interactions with other vehicles on the road. Since the details of these interactions emerge in the moment and cannot be known in advance, precisely predicting what happens to the the car is possible only a few minutes into the future.

AI is already playing an enormous role in health care.

Not with my health

These same principles apply to prescribing medications. Different conditions and diseases can have the same symptoms, and people with the same condition or disease may exhibit different symptoms. For example, fever can be caused by a respiratory illness or a digestive one. And a cold might cause cough, but not always.

This means that health care datasets have significant overlaps that would prevent AI from being error-free.

Certainly, humans also make errors. But when AI misdiagnoses a patient, as it surely will, the situation falls into a legal limbo. It’s not clear who or what would be responsible if a patient were hurt. Pharmaceutical companies? Software developers? Insurance agencies? Pharmacies?

In many contexts, neither humans nor machines are the best option for a given task. “Centaurs,” or “hybrid intelligence” – that is, a combination of humans and machines – tend to be better than each on their own. A doctor could certainly use AI to decide potential drugs to use for different patients, depending on their medical history, physiological details and genetic makeup. Researchers are already exploring this approach in precision medicine.

But common sense and the precautionary principle suggest that it is too early for AI to prescribe drugs without human oversight. And the fact that mistakes may be baked into the technology could mean that where human health is at stake, human supervision will always be necessary.

Carlos Gershenson, Professor of Innovation, Binghamton University, State University of New York

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read next:

• If social media for kids is so bad, should we be allowed to post kids’ photos online?

• Meta Reminds Users That AI Interactions Will Soon Shape Content and Ads

• AI And Marketing: How Connecting With Customers In The Real World Gives Brands The Advantage

• OpenAI Flags High Cybersecurity Risks in Future AI Models

by Web Desk via Digital Information World

If social media for kids is so bad, should we be allowed to post kids’ photos online?

As Australia’s ban on under-16-year-olds having certain social media accounts kicks in this week, debate on whether it’s a good idea or even legal rages on – both at home and overseas.

Image: Samsung Memory / Unsplash

Yet barely acknowledged in this debate is what happens when a child doesn’t have an account, yet their entire childhood is still documented online. Should this be permitted?

“Sharenting” – when parents share their children’s lives online – entered the dictionary a few years ago. Awareness of potential risks has been increasing, but many parents still routinely share pictures and videos of their children online.

Sharenting is widespread and persistent. A review of practices over the past ten years describes that parents commonly share details such as children’s names, dates of birth, birthday parties, milestones (birthdays, school achievements), health info and photos. This produces a “digital identity” of the child long before they can consent.

And it’s not just parents. Dance schools, soccer clubs and various other community groups, as well as family members and friends, commonly post about children online. All contribute to what’s essentially a collective digital album about the child. Even for children not yet old enough to have their own account, their lives could be heavily documented online until they do.

This challenge moves us well beyond traditional approaches to safety messages such as “don’t share your personal details online” or “don’t talk to strangers”. It requires a deeper understanding of what exactly safety and wellbeing for children on online platforms looks like.

A passive data subject

Here’s a typical sharenting scenario. A family member uploads a photo captioned “Mia’s 8th birthday at Bondi beach!” to social media, where it gets tagged and flooded with comments from relatives and friends.

Young Mia isn’t scrolling. She isn’t being bullied. She doesn’t have her own account. But in the act of having a photo and multiple comments about her uploaded, she has just become a passive data subject. Voluntarily disclosed by others, Mia’s sensitive information – data on her face and age – exposes her to risks without her consent or participation.

The algorithm doesn’t care Mia is eight years old. It cares that her photo keeps adults on the app for longer. Her digital persona is being used to sustain the platform’s real product: adult attention. Children’s images posted by family and friends function as engagement tools, with parents reporting that “likes” and comments encourage them to continue sharing more about their child.

We share such posts to connect with family and to feel part of a community. Yet a recent Italian study of 228 parents found 93% don’t fully realise the associated data harvesting practices that take place, and their risk to the child’s privacy, security and image protection.

A public narrative of one’s life

Every upload of a child’s face, especially across years and from multiple sources, help create a digital identity they don’t have control over. Legally and ethically, many frameworks attempt to restrict commercial data profiling of minors, but recent studies show profiling is still happening at scale.

By the time a child is 16 – old enough to create their own account – a platform may already have accumulated a sizeable and lucrative profile of them to sell to advertisers.

The fallout isn’t just about data; it’s personal. That cute birthday photo can resurface in a background check for future employment or become ammunition for teenage bullying.

More subtly, a young person forging their identity must now contend with a pre-written, public narrative of their life, one they didn’t choose or control.

New laws aiming to ban children from social media address real harms such as exposure to misogynistic or hateful material, dangerous online challenges, violent videos, and content promoting disordered eating and suicide – but they focus on the child as a user. In today’s data economy, you don’t need an account to be tracked and profiled. You just need to be relevant to someone else who has an account.

What can we do?

The essential next step is social media literacy for all of us. This is a new form of literacy for the digital world we live in now. It means understanding how algorithms shape our feeds, how dark design patterns keep us scrolling, and that any “like” or photo is a data point in a vast commercial machine.

Social media literacy is not just for kids in classrooms, but for parents, coaches, carers and anyone else engaging with kids in our online world. We all need to understand this.

Sharenting-awareness campaigns exist, from eSafety’s parental privacy resources, to the EU-funded children’s digital rights initiative, but they are not yet shifting the culture. That’s because we’re conditioned to think about our children’s physical safety, not so much their data safety. Because the risks of posting aren’t immediate or visible, its easy to underestimate them.

Shifting adult behaviour closes the gap between our concerns and our actions, and the reality of children’s exposure to content on social media.

Keeping children safe online means looking beyond kids as users and recognising the role adults play in creating a child’s digital footprint.

Joanne Orlando, Researcher, Digital Wellbeing, Western Sydney University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read next: AI And Marketing: How Connecting With Customers In The Real World Gives Brands The Advantage

by Web Desk via Digital Information World

OpenAI Flags High Cybersecurity Risks in Future AI Models

OpenAI (the maker of ChatGPT) expects its upcoming artificial intelligence (AI) models to potentially reach high levels of cybersecurity capability, potentially enabling the exploitation of previously unknown software vulnerabilities or assisting in complex enterprise or industrial intrusions. The company emphasized that these risks are part of broader dual-use capabilities that may also benefit defenders.

To address these concerns, OpenAI is investing in defensive measures, including tools for auditing code, patching vulnerabilities, and supporting security workflows. The company is implementing a layered, defense-in-depth approach including "access controls, infrastructure hardening, egress controls," along with monitoring, detection systems, and threat intelligence programs.

OpenAI plans to introduce a trusted access program allowing qualified cyberdefense professionals to employ advanced capabilities for defensive purposes. It is also establishing the Frontier Risk Council, an advisory group of cybersecurity professionals to guide safe deployment and evaluate potential misuse.

Additional initiatives include Aardvark, an agentic security researcher currently in private beta designed to identify and help patch software vulnerabilities, and collaboration with the Frontier Model Forum to develop shared understanding of threat models across the AI industry. OpenAI frames these measures as ongoing, long-term investments to strengthen defenses and mitigate risks associated with increasingly capable AI systems.

Notes: This post was drafted with the assistance of AI tools and reviewed, edited, and published by humans. Image: DIW-AIgen

Read next: Google Expands Android’s Safety Features With Emergency Live Video Rollout
by Ayaz Khan via Digital Information World

Google Expands Android’s Safety Features With Emergency Live Video Rollout

Google has begun rolling out Emergency Live Video on Android, introducing a way for users to share real-time visual information with emergency responders during calls or texts. The feature allows dispatchers to send a request to a user’s device when they determine that viewing the scene would help them assess the situation and provide timely assistance.

Users receive an on-screen prompt and can choose whether to share their camera feed. The stream is encrypted by default, and users retain full control throughout the process, with the ability to stop transmission instantly. The feature requires no setup and is designed to operate through a single, direct action on the user’s device.

Emergency Live Video is intended to support responders in evaluating incidents such as medical crises or fast-moving hazards, and it can help them guide callers through urgent steps until aid arrives. The capability expands Google’s existing emergency-focused tools, including Emergency Location Service, Car Crash Detection, Fall Detection and Satellite SOS.

The rollout begins across the United States and select regions in Germany and Mexico. Devices running Android 8 or later with Google Play services support the feature. Google is working with public safety organizations worldwide to extend availability, and interested agencies can access partner documentation.

Notes: This post was drafted with the assistance of AI tools and reviewed, edited, and published by humans.

Read next: Studies Reveal Severe Gen Z Burnout and Recommend Stronger Workplace Support and Clearer Expectations
by Asim BN via Digital Information World

Studies Reveal Severe Gen Z Burnout and Recommend Stronger Workplace Support and Clearer Expectations

Gen Z workers are reporting some of the highest burnout levels ever recorded, with new research suggesting they are buckling under unprecedented levels of stress.

While people of all age levels report burnout, Gen Z and millennials are reporting “peak burnout” at earlier ages. In the United States, a poll of 2,000 adults found that a quarter of Americans are burnt out before they’re 30 years old.

Image: Vitaly Gariev / Unsplash

Similarly, a British study measured burnout over an 18-month period after the COVID-19 pandemic and found Gen Z members were reporting burnout levels of 80 per cent. Higher levels of burnout among the Gen Z cohort were also reported by the BBC a few years ago.

Globally, a survey covering 11 countries and more than 13,000 front-line employees and managers reported that Gen Z workers were more likely to feel burnt out (83 per cent) than other employees (75 per cent).

Another international well-being study found that nearly one-quarter of 18- to 24-year-olds were experiencing “unmanageable stress,” with 98 per cent reporting at least one symptom of burnout.

And in Canada, a Canadian Business survey found that 51 per cent of Gen Z respondents felt burnt out — lower than millennials at 55 per cent, but higher than boomers at 29 per cent and Gen X, at 32 per cent.

As a longstanding university educator of Gen Z students, and a father of two of this generation, the levels of Gen Z burnout in today’s workplace are astounding. Rather than dismissing young workers as distracted or too demanding of work-life balance, we might consider that they’re sounding the alarm of what’s broken at work and how we can fix it.

What burnout really is

Burnout can vary from person to person and across occupations, but researchers generally agree on its core features. It occurs when there is conflict between what a worker expects from their job and what the job actually demands.

That mismatch can take many forms: ambiguous job tasks, an overload of tasks or not having enough resources or the skills needed to respond to a role’s demands.

In short, burnout is more likely to occur when there’s a growing mismatch between one’s expectations of work and its actual realities. Younger workers, women and employees with less seniority are consistently at higher risk of burnout.

Burnout typically progresses across three dimensions. While fatigue is often the first noticeable symptom of burnout, the second is cynicism or depersonalization, which leads to alienation and detachment to one’s work. This detachment leads to the third dimension of burnout: a declining sense of personal accomplishment or self-efficacy.

• Also read: Research Tracks 8,324 U.S. Children, Identifying Social Media as a Risk Factor for Growing Inattention

Why Gen Z is especially vulnerable to burnout

Several forces converge to make Gen Z particularly susceptible to burnout. First, many Gen Z entered the workforce during and after the COVID-19 pandemic.

It was a time of profound upheaval, social isolation and changing work protocols and demands. These conditions disrupted the informal learning that typically happens through everyday interactions with colleagues that were hard to replicate in a remote workforce.

Second, broader economic pressures have intensified. As American economist Pavlina Tcherneva argues, the “death of the social contract and the enshittification of jobs” — the expectation that a university education would result in a well-paying job — have left many young people navigating a far more precarious landscape.

The intensification of economic disruption, widening inequality, increasing costs of housing and living and the rise of precarious employment have put greater financial pressures on this generation.

A third factor is the restructuring of work that is taking place under artificial intelligence. As workplace strategist Ann Kowal Smith wrote in a recent Forbes article, Gen Z is the first generation to enter a labour market defined by a “new architecture of work: hybrid schedules that fragment connection, automation that strips away context and leaders too busy to model judgment.”

What can be done?

If you’re reading this and feeling burnt out, the first thing to know is that you’re not overreacting and you’re not alone. The good news is, there are ways to recover.

One of burnout’s most overlooked antidotes is combating the alienation and isolation it produces. The best way to do this is by building connection and relation to others, starting with work colleagues. This could be as simple as checking in with a teammate after a meeting or setting up a weekly coffee with a colleague.

In addition, it’s important to give up on the idea that excessive work is better work. Set boundaries at work by blocking out time in your calendar and clearly signalling your availability to colleagues.

But individual coping strategies can only go so far. The more fundamental solutions must come from workplaces themselves. Employers need to offer more flexible work arrangements, including wellness and mental health supports. Leaders and managers should communicate job expectations clearly, and workplaces should have policies to proactively review and redistribute excessive workloads.

Kowal Smith has also suggested building a new “architecture of learning” in the workplace that includes mentorship, provides feedback loops and rewards curiosity and agility.

Taken together, these workplace transformation efforts could humanize the workplace, lessen burnout and improve engagement, even at a time of encroaching AI. A workplace that works better for Gen Z ultimately works better for all of us.

Nitin Deckha, Lecturer in Justice Studies, Early Childhood Studies, Community and Social Services and Electives, University of Guelph-Humber

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read next: Pew Survey: 64% of Teens Use AI Chatbots, and 97% Go Online Daily

by Web Desk via Digital Information World

Research Tracks 8,324 U.S. Children, Identifying Social Media as a Risk Factor for Growing Inattention

A longitudinal study, published in Pediatrics Open Science, following 8,324 children aged 9 to 14 in the United States has found that social media use is associated with a gradual increase in inattention symptoms. Researchers at Karolinska Institutet in Sweden and Oregon Health & Science University tracked children annually for four years, assessing time spent on social media, television/videos, and video games alongside parent-reported attention measures.

On average, children spent 2.3 hours per day watching television or videos, 1.5 hours on video games, and 1.4 hours on social media. Only social media use was linked to growing inattention over time. The effect was small for individual children but could have broader consequences at the population level. Hyperactivity and impulsive behaviors were not affected.

The association remained consistent regardless of sex, ADHD diagnosis, genetic predisposition, socioeconomic status, or ADHD medication. Children with pre-existing inattention symptoms did not increase their social media use, indicating the relationship primarily runs from use to symptoms.

Researchers note that social media platforms can create mental distractions through notifications and messages, potentially reducing the ability to focus. The study does not suggest all children will experience attention difficulties but highlights the importance of informed decisions regarding digital media exposure.

The research team plans to continue monitoring the participants beyond age 14. The study was funded by the Swedish Research Council and the Masonic Home for Children in Stockholm, with no reported conflicts of interest.

Source: “Digital Media, Genetics and Risk for ADHD Symptoms in Children – A Longitudinal Study,” Pediatrics Open Science, 2025.

Notes: This post was drafted with the assistance of AI tools and reviewed, edited, and published by humans.

Image: Vikas Makwana / unsplash

Read next: Pew Survey: 64% of Teens Use AI Chatbots, and 97% Go Online Daily
by Asim BN via Digital Information World