Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

A new study analysed 1.5 million conversations with the AI chatbot Claude, revealing rare but concerning cases of what some are calling "AI psychosis". Sky's Rowland Manthorpe reports

Sky News

8,881,844 subscribers

26,639 views • 3 months ago •via X (Twitter)

News & Politics Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

John Ternus has been announced as the new CEO of Apple, but how will he cope with his predecessor Tim Cook's big AI bet? Sky's Rowland Manthorpe explores the big challenge facing the new Apple CEO

John Ternus has been announced as the new CEO of Apple, but how will he cope with his predecessor Tim Cook's big AI bet? Sky's Rowland Manthorpe explores the big challenge facing the new Apple CEO

Sky News

28,131 views • 2 months ago

AI researchers are resigning - what does that mean? Media reports and social media posts have described the resignations as a “wave”. On closer inspection, however, the moves have little in common. Sky's Tech correspondent Rowland Manthorpe explains.

AI researchers are resigning - what does that mean? Media reports and social media posts have described the resignations as a “wave”. On closer inspection, however, the moves have little in common. Sky's Tech correspondent Rowland Manthorpe explains.

Sky News

23,030 views • 5 months ago

When leading AI company Anthropic launched its latest AI model, Claude Opus 4.6, at the end of last week, it broke many measures of intelligence and effectiveness - including one crucial benchmark: the vending machine test. Sky's Rowland Manthorpe explains the experiment

When leading AI company Anthropic launched its latest AI model, Claude Opus 4.6, at the end of last week, it broke many measures of intelligence and effectiveness - including one crucial benchmark: the vending machine test. Sky's Rowland Manthorpe explains the experiment

Sky News

36,406 views • 5 months ago

Artificial intelligence can scan vast amounts of data and use it to flag targets, rank threats and suggest priorities - but what are the dangers? Sky's technology correspondent Rowland Manthorpe on AI, warfare and the dangers within ⬇️ Read more 🔗

Artificial intelligence can scan vast amounts of data and use it to flag targets, rank threats and suggest priorities - but what are the dangers? Sky's technology correspondent Rowland Manthorpe on AI, warfare and the dangers within ⬇️ Read more 🔗

Sky News

22,730 views • 4 months ago

"AI cope is classifying anybody who has a positive experience with AI as being in AI psychosis." Marc Andreessen 🇺🇸 explains AI cope vs. AI psychosis: " You tell Claude you've discovered a new idea for an anti-gravity machine. And Claude says, 'Oh, that's amazing... Like, you've achieved a giant breakthrough in physics.'" "Yes there's some number of those cases." "But that causes kind of AI critics or AI doomers to basically say anybody therefore who reports a positive, productive experience with AI has fallen into AI psychosis." "And so anybody who actually is like, 'Wow, my productivity is way up,' or, 'Wow, I really have a thought partner for the first time in my work,' or, 'Wow, I really have been able to produce something that I never would have been able produced before,' that's sort of all bucketed under they all have AI psychosis."

"AI cope is classifying anybody who has a positive experience with AI as being in AI psychosis." Marc Andreessen 🇺🇸 explains AI cope vs. AI psychosis: " You tell Claude you've discovered a new idea for an anti-gravity machine. And Claude says, 'Oh, that's amazing... Like, you've achieved a giant breakthrough in physics.'" "Yes there's some number of those cases." "But that causes kind of AI critics or AI doomers to basically say anybody therefore who reports a positive, productive experience with AI has fallen into AI psychosis." "And so anybody who actually is like, 'Wow, my productivity is way up,' or, 'Wow, I really have a thought partner for the first time in my work,' or, 'Wow, I really have been able to produce something that I never would have been able produced before,' that's sort of all bucketed under they all have AI psychosis."

MTS

10,801 views • 2 months ago

AI is getting much better at hacking and fast. Data from the UK AI Security Institute, the Center for AI Standards and Innovations and others, shows how AI models are becoming dramatically more capable at cyberattacks. Sky's Rowland Manthorpe explains why this is no longer just a warning about one powerful model, but a much bigger industry-wide shift. 📺

AI is getting much better at hacking and fast. Data from the UK AI Security Institute, the Center for AI Standards and Innovations and others, shows how AI models are becoming dramatically more capable at cyberattacks. Sky's Rowland Manthorpe explains why this is no longer just a warning about one powerful model, but a much bigger industry-wide shift. 📺

Sky News

18,021 views • 1 month ago

The scribes of medieval Europe and AI coding: Is there a connection between the two? Anthropic's Head of Claude Coding seems to think so. Technology correspondent Rowland Manthorpe asked Boris Cherny to explain. More science and tech stories:

The scribes of medieval Europe and AI coding: Is there a connection between the two? Anthropic's Head of Claude Coding seems to think so. Technology correspondent Rowland Manthorpe asked Boris Cherny to explain. More science and tech stories:

Sky News

17,144 views • 1 month ago

they stole Claude's brain 16 million times Anthropic just exposed DeepSeek, Moonshot AI, and MiniMax for creating 24,000 fake accounts and having 16 million conversations with Claude to steal its capabilities. The AI race isn't what you think. Source: #ai #claude #deepseek #cybersecurity #espionage #anthropic

they stole Claude's brain 16 million times Anthropic just exposed DeepSeek, Moonshot AI, and MiniMax for creating 24,000 fake accounts and having 16 million conversations with Claude to steal its capabilities. The AI race isn't what you think. Source: #ai #claude #deepseek #cybersecurity #espionage #anthropic

NetworkChuck

22,886 views • 4 months ago

Satellites and AI are transforming how the world is watched, connected and targeted - and the US, along with Elon Musk’s SpaceX, is at the centre of that shift ⁠ Rowland Manthorpe investigates the rise of a planetary watch machine and a “digital kill web”

Satellites and AI are transforming how the world is watched, connected and targeted - and the US, along with Elon Musk’s SpaceX, is at the centre of that shift ⁠ Rowland Manthorpe investigates the rise of a planetary watch machine and a “digital kill web”

Sky News

212,870 views • 1 month ago

Sky News technology correspondent Rowland Manthorpe has spent the last few months investigating AI music. What has emerged is a picture of a vast attempted fraud, as technologically-equipped criminals use AI tools to try to take billions of pounds away from real-life musicians.

Sky News technology correspondent Rowland Manthorpe has spent the last few months investigating AI music. What has emerged is a picture of a vast attempted fraud, as technologically-equipped criminals use AI tools to try to take billions of pounds away from real-life musicians.

Sky News

18,625 views • 5 months ago

Today there is a new danger on the toy market—AI. This harmless looking teddy bear actually has an AI chatbot embedded inside. With this tech kids can engage in conversations with their toys that are intimate, inappropriate & deeply dangerous.

Today there is a new danger on the toy market—AI. This harmless looking teddy bear actually has an AI chatbot embedded inside. With this tech kids can engage in conversations with their toys that are intimate, inappropriate & deeply dangerous.

Richard Blumenthal

30,041 views • 7 months ago

.Keith Sakata, MD (VC Fellow, Scrub Capital) on AI and mental health risks He says AI isn’t directly causing psychosis, but it can intensify existing vulnerabilities. "AI isn't causing psychosis by itself, but it can supercharge vulnerabilities. Psychosis really thrives when reality stops pushing back, and AI can soften that wall for some people." "For example, AI can accelerate or augment the severity of underlying issues, creating a feedback loop." Sakata also clarified that “AI psychosis is not a clinical term. Psychosis is a symptom, not a diagnosis.”

.Keith Sakata, MD (VC Fellow, Scrub Capital) on AI and mental health risks He says AI isn’t directly causing psychosis, but it can intensify existing vulnerabilities. "AI isn't causing psychosis by itself, but it can supercharge vulnerabilities. Psychosis really thrives when reality stops pushing back, and AI can soften that wall for some people." "For example, AI can accelerate or augment the severity of underlying issues, creating a feedback loop." Sakata also clarified that “AI psychosis is not a clinical term. Psychosis is a symptom, not a diagnosis.”

TBPN

26,323 views • 11 months ago

Cases of shoplifting have more than doubled in the past five years in England and Wales, topping more than half a million cases a year, according to House of Commons Library data analysed by the Liberal Democrats. The data also shows that most shoplifters are getting away with their crime, with charges only being brought in as few as one in every 14 cases in some areas. Nick Dixon reports.

Cases of shoplifting have more than doubled in the past five years in England and Wales, topping more than half a million cases a year, according to House of Commons Library data analysed by the Liberal Democrats. The data also shows that most shoplifters are getting away with their crime, with charges only being brought in as few as one in every 14 cases in some areas. Nick Dixon reports.

Good Morning Britain

11,129 views • 3 months ago

Poe is one of the most underrated AI tools right now. The Preview feature allows you to code custom apps directly in chat, similar to Claude Artifacts, but for ANY AI model (GPT-4o, Gemini 1.5, + more) Here's a few of my favorite use cases:

Poe is one of the most underrated AI tools right now. The Preview feature allows you to code custom apps directly in chat, similar to Claude Artifacts, but for ANY AI model (GPT-4o, Gemini 1.5, + more) Here's a few of my favorite use cases:

Rowan Cheung

155,657 views • 2 years ago

The new Chinese AI-chatbot DeepSeek caused US tech firms a loss of over $1 trillion - but why?

The new Chinese AI-chatbot DeepSeek caused US tech firms a loss of over $1 trillion - but why?

Channel 4 News

13,592 views • 1 year ago

Saving a million lives a year with AI sounds wonderful. But what are the hidden dangers of this #AlienIntelligence, and how hard are we working to figure out #AI #threats? Full conversation with richroll from the richrollPodcast:

Saving a million lives a year with AI sounds wonderful. But what are the hidden dangers of this #AlienIntelligence, and how hard are we working to figure out #AI #threats? Full conversation with richroll from the richrollPodcast:

Yuval Noah Harari

27,176 views • 1 year ago

You can now build apps for Meta’s AI Glasses with the new Wearables Device Access Toolkit! Here are some of the first use cases. Very cool.

You can now build apps for Meta’s AI Glasses with the new Wearables Device Access Toolkit! Here are some of the first use cases. Very cool.

Justin Ryan

29,004 views • 10 months ago

Thousands of Serbians are marching across the country to honour the 16 victims of the Novi Sad rail disaster. But now they are also calling for new elections and democratic reform. Sky's Alistair Bunkall reports

Thousands of Serbians are marching across the country to honour the 16 victims of the Novi Sad rail disaster. But now they are also calling for new elections and democratic reform. Sky's Alistair Bunkall reports

Sky News

92,683 views • 8 months ago

Something interesting is happening in AI right now. 1.5 million people are leaving ChatGPT. Not because AI is slowing down. Because more tools are appearing every week. People are testing new platforms for: → Different features. → Faster responses. → Specialized workflows. → Different integrations. The AI space is evolving fast. And users are experimenting with what works best.

Something interesting is happening in AI right now. 1.5 million people are leaving ChatGPT. Not because AI is slowing down. Because more tools are appearing every week. People are testing new platforms for: → Different features. → Faster responses. → Specialized workflows. → Different integrations. The AI space is evolving fast. And users are experimenting with what works best.

Julian Goldie SEO

10,151 views • 4 months ago