Artificial intelligence (AI) giant Anthropic has expanded access to a highly advanced model deemed too dangerous for public release, including Australia in the select handful of users.

The large language model, known as Claude Mythos, is now being rolled out to an additional 150 organisations across 15 countries, including the Australian government and several local businesses, as part of Project Glasswing.

In an era where large-scale AI launches are happening on a day-by-day basis, this limited, gradual release may seem particularly surprising. But Mythos is not like most other AI systems. Instead it’s an automated tool for assessing software to find critical bugs and vulnerabilities.

This managed release is deliberate, as the discovery of vulnerabilities in computer systems is useful for those who want to defend them and those who want to hack them.

However, the real nature of the impact of AI systems on cybersecurity is significantly more complex.

Finding hundreds of severe vulnerabilities

Under initial testing, Mythos has been able to identify multiple new high-risk vulnerabilities. Left unfixed, such flaws allow attackers to easily steal data or induce system crashes.

While these reports are promising, the raw data needs context. Of the 23,000 vulnerabilities flagged by Mythos, only 6,200 were estimated as high-risk by Mythos. However AI isn’t perfect, as human experts could only validate two in every three of these vulnerabilities as high-risk. Even still, the nature and severity of identified vulnerabilities has led developers to say that with Mythos “defenders finally have a chance to win, decisively”.

And winning this battle is extremely valuable.

Over the last few years, Australians have repeatedly been the victims of costly cybersecurity incidents, including Optus, Medibank Private, the Melbourne International Film Festival, and Canvas.

This barrage of attacks likely explain why the Australian Signals Directorate welcomed Australia’s inclusion in Anthropic’s Project Glasswing. While this AI-driven security offers huge potential benefits, the government so far has been tight-lipped on the specifics of how Mythos will actually be used.

Dangerous in the wrong hands

While discovering vulnerabilities is useful, defenders need to be able to respond to them. This is problematic when tools like Mythos produce large numbers of false reports, which have the potential to overwhelm unprepared cybersecurity teams.

More concerningly, while access to Mythos is currently tightly controlled, it will not be long until similar tools are available to help support hackers.

And it’s not just the vulnerabilities that AI can discover that pose risks.

AI systems more broadly are incredibly vulnerable to being tricked or exploited, with highly damaging consequences.

Just this week, hackers used Meta’s AI powered chatbot to gain access to high-profile Instagram accounts, including Barack Obama’s. They did so by tricking AI chatbots into changing account details. And, even after Instagram announced it fixed the issue, within hours there were reports of further accounts being compromised.

A similar attack known as Echoleak last year revealed how tying Microsoft Copilot to email accounts could introduce significant risks. This was made possible by sending emails to accounts monitored by Copilot’s AI. These emails tricked the AI into leaking large amounts of private and confidential information, without the email ever needing to be opened by a human. No longer do we live in a world where hackers need to convince users to click a malicious link, if they can instead convince the AI that reads emails to act dangerously.

Both Echoleak and the Instagram hacks underscore the risks we face as more and more organisations tie their critical functions to AI systems that are difficult to audit, and easy to exploit – even by just being persuasive.

A new balance point

All of this suggests the current cybersecurity landscape might be shifting to a new balance point, where defenders and hackers race to develop and exploit powerful AI tools.

Tools like Mythos aren’t a silver bullet. While they provide defenders with an additional set of eyes on where to look, it still will require expertise to work out what is real, and what isn’t.

But the advent of the AI era has already fundamentally changed the risks associated with poor cybersecurity practices. Every day a user or service provider delays a software update on one of their devices is a day where a vulnerability can be exploited.

For cybersecurity teams, ensuring compliance is already a difficult enough process that will only get worse when the speed of vulnerability discovery increases.

While they are high value targets for hackers, large organisations will likely remain safe, as they will have the resources to access and deploy tools like Mythos. But smaller, less resourced companies will likely not have the capacity to access these tools – or to react to the upcoming tsunami of cybersecurity updates.

And if they fall behind on these updates, these smaller companies will likely find themselves at far more risk than they ever have been before.

The cybersecurity divide between those with and without resources will only grow. Bridging this gap is not just an IT challenge – it’s a public safety concern that will affect us all.

Andrew Cullen does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

More people are relying on social media – such as TikTok, Instagram, YouTube and Reddit – to learn about mental health conditions and to interact with people who have shared experiences.

These aren’t only long-familiar disorders such as depression, anxiety and schizophrenia. They also include conditions often placed under the “neurodivergent” umbrella such as autism, ADHD (attention-deficit hyperactivity disorder), Tourette syndrome and dyslexia.

For instance, on TikTok the hashtag #adhd has had more than 50 billion views.

We wanted to explore how social media platforms shape how we understand mental health. So we analysed more than 14 million posts and comments about mental health on Reddit.

We show a shift in conversations toward ADHD and autism, and away from anxiety and depression.

Our findings have important implications for how people make sense of, and seek help for, mental health problems.

A complex relationship

Social media coverage of mental health has made it more visible, with some positive effects. It has probably reduced the stigma of mental illness and increased the use of mental health services.

However, it also has downsides. It can induce or exacerbate eating disorders, can contribute to the spread of symptoms (such as tic-like behaviours), and has been attributed to the rise of questionable self-diagnoses.

Misinformation is common in social media discussions of mental health. One study found a majority of the most popular TikTok videos on ADHD were misleading. Inaccurate information about many other mental health conditions on social media is common.

Discussions change and evolve

Mental health content has not merely risen in volume. Some conditions have increasingly attracted the spotlight, others have receded from view, and the relationships among them have shifted.

In our Reddit study published last year, we found that as the largest ADHD- and autism- related communities (subreddits) became increasingly more prominent from 2012 to 2022, their content gradually became more similar, and their users increasingly overlapped.

Discussions in both communities increasingly emphasised the experiences of adults, challenges in accessing diagnostic assessments, and struggles with personal relationships.

This growing convergence of these two conditions on Reddit illustrates how social media can reshape representations of mental health.

Our latest study takes this further

In our new study, we analysed more than 14 million posts and comments from several of the largest mental health communities on Reddit.

The 14 communities we studied included those related to mood, anxiety, trauma, personality, dissociation and psychosis, as well as those focused on conditions often placed under the “neurodivergent” umbrella, such as autism, ADHD, Tourette syndrome and dyslexia.

We investigated how the people belonging to these communities and the language they used changed from 2015 to 2022.

We explored which communities became more or less closely associated over time – sharing more or fewer members and containing posts and comments with similar or different linguistic content. We also looked at whether these changes reflected shifts in the amount of attention the 14 conditions received.

Although our analysis only covered a seven-year period, it revealed a striking pattern of changes. The two diagrams show how the 14 communities were interrelated at the beginning and end of the period.

The size of the circles represents the relative size of the communities. The width of the links between them indicates how closely they were associated.

In 2015, depression and anxiety were prominent mental health communities on Reddit. They were among the most active and their members and content overlapped with those of many other communities. In this sense, they were “central” to the network.

However, in 2022, ADHD and autism communities had become most popular and prominent, displacing depression and anxiety. ADHD, autism and other neurodivergent conditions became more closely associated with other communities, and consequently more central to the network.

These analyses suggest that on Reddit the mental health landscape has been re-configured. Mood and anxiety disorders once dominated discussions. But discussions of mental health have increasingly pivoted to discussing conditions related to being neurodivergent.

Reddit users do not represent the general population; they tend to be younger, male, more educated, and have a higher income. Nevertheless, our study offers important insights into changes in mental health discussions on one social media platform over time.

Why does it matter?

The rising prominence and centrality of ADHD and autism makes them increasingly popular explanations for mental health problems. This might promote accurate self-diagnosis by people who once would not have recognised the nature of their difficulties.

However, it could also lead people to misinterpret and mislabel their experiences as ADHD and autism when there’s another explanation.

The rising prominence of these conditions on social media may also lead people to interpret mood or anxiety symptoms as signs of ADHD or autism.

Misinterpretations can lead people to pursue inappropriate diagnoses or unhelpful treatment, delaying access to the help they need. This in turn places increasing pressure on mental health services, and can lead to other conditions being overlooked.

Jemima Kang receives funding from an Australian Government Research Training Program Scholarship, an Australian Academy of Technological Sciences and Engineering Elevate Scholarship, and a University of Melbourne Helen Macpherson Smith Scholarship.

Nick Haslam receives funding from the Australian Research Council.

Mike Conway does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

More people are relying on social media – such as TikTok, Instagram, YouTube and Reddit – to learn about mental health conditions and to interact with people who have shared experiences.

These aren’t only long-familiar disorders such as depression, anxiety and schizophrenia. They also include conditions often placed under the “neurodivergent” umbrella such as autism, ADHD (attention-deficit hyperactivity disorder), Tourette syndrome and dyslexia.

For instance, on TikTok the hashtag #adhd has had more than 50 billion views.

We wanted to explore how social media platforms shape how we understand mental health. So we analysed more than 14 million posts and comments about mental health on Reddit.

We show a shift in conversations toward ADHD and autism, and away from anxiety and depression.

Our findings have important implications for how people make sense of, and seek help for, mental health problems.

A complex relationship

Social media coverage of mental health has made it more visible, with some positive effects. It has probably reduced the stigma of mental illness and increased the use of mental health services.

However, it also has downsides. It can induce or exacerbate eating disorders, can contribute to the spread of symptoms (such as tic-like behaviours), and has been attributed to the rise of questionable self-diagnoses.

Misinformation is common in social media discussions of mental health. One study found a majority of the most popular TikTok videos on ADHD were misleading. Inaccurate information about many other mental health conditions on social media is common.

Discussions change and evolve

Mental health content has not merely risen in volume. Some conditions have increasingly attracted the spotlight, others have receded from view, and the relationships among them have shifted.

In our Reddit study published last year, we found that as the largest ADHD- and autism- related communities (subreddits) became increasingly more prominent from 2012 to 2022, their content gradually became more similar, and their users increasingly overlapped.

Discussions in both communities increasingly emphasised the experiences of adults, challenges in accessing diagnostic assessments, and struggles with personal relationships.

This growing convergence of these two conditions on Reddit illustrates how social media can reshape representations of mental health.

Our latest study takes this further

In our new study, we analysed more than 14 million posts and comments from several of the largest mental health communities on Reddit.

The 14 communities we studied included those related to mood, anxiety, trauma, personality, dissociation and psychosis, as well as those focused on conditions often placed under the “neurodivergent” umbrella, such as autism, ADHD, Tourette syndrome and dyslexia.

We investigated how the people belonging to these communities and the language they used changed from 2015 to 2022.

We explored which communities became more or less closely associated over time – sharing more or fewer members and containing posts and comments with similar or different linguistic content. We also looked at whether these changes reflected shifts in the amount of attention the 14 conditions received.

Although our analysis only covered a seven-year period, it revealed a striking pattern of changes. The two diagrams show how the 14 communities were interrelated at the beginning and end of the period.

The size of the circles represents the relative size of the communities. The width of the links between them indicates how closely they were associated.

In 2015, depression and anxiety were prominent mental health communities on Reddit. They were among the most active and their members and content overlapped with those of many other communities. In this sense, they were “central” to the network.

However, in 2022, ADHD and autism communities had become most popular and prominent, displacing depression and anxiety. ADHD, autism and other neurodivergent conditions became more closely associated with other communities, and consequently more central to the network.

These analyses suggest that on Reddit the mental health landscape has been re-configured. Mood and anxiety disorders once dominated discussions. But discussions of mental health have increasingly pivoted to discussing conditions related to being neurodivergent.

Reddit users do not represent the general population; they tend to be younger, male, more educated, and have a higher income. Nevertheless, our study offers important insights into changes in mental health discussions on one social media platform over time.

Why does it matter?

The rising prominence and centrality of ADHD and autism makes them increasingly popular explanations for mental health problems. This might promote accurate self-diagnosis by people who once would not have recognised the nature of their difficulties.

However, it could also lead people to misinterpret and mislabel their experiences as ADHD and autism when there’s another explanation.

The rising prominence of these conditions on social media may also lead people to interpret mood or anxiety symptoms as signs of ADHD or autism.

Misinterpretations can lead people to pursue inappropriate diagnoses or unhelpful treatment, delaying access to the help they need. This in turn places increasing pressure on mental health services, and can lead to other conditions being overlooked.

Jemima Kang receives funding from an Australian Government Research Training Program Scholarship, an Australian Academy of Technological Sciences and Engineering Elevate Scholarship, and a University of Melbourne Helen Macpherson Smith Scholarship.

Nick Haslam receives funding from the Australian Research Council.

Mike Conway does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.