Close Menu
  • Home
  • IRAQ
  • MIDDLE EAST
  • WORLD
  • Business
  • Lifestyle
  • Sports
  • Tech
  • More
    • Media & Culture
    • Health
    • Iraqis in Europe

Subscribe to Updates

Subscribe Us To Receive Our Latest News Directly In Your Inbox!

We don’t spam! Read our privacy policy for more info.

Check your inbox or spam folder to confirm your subscription.

What's Hot

Hannah Gutierrez-Reed: Weapons supervisor convicted in fatal shooting on Alec Baldwin film set freed from jail | Ents & Arts News

May 25, 2025

Inside Kering’s 10-year partnership with Cannes Film Festival

May 25, 2025

Over de Liefde: ‘Door met mezelf in relatietherapie te gaan ben ik een leukere partner geworden’

May 25, 2025
Facebook X (Twitter) Instagram
Trending
  • Hannah Gutierrez-Reed: Weapons supervisor convicted in fatal shooting on Alec Baldwin film set freed from jail | Ents & Arts News
  • Inside Kering’s 10-year partnership with Cannes Film Festival
  • Over de Liefde: ‘Door met mezelf in relatietherapie te gaan ben ik een leukere partner geworden’
  • ‘I want a child but I’m scared to come off the pill’
  • Gail’s backer plots rare move with bid for steak chain Flat Iron | Money News
  • US lifts first sanctions on Syria following Trump’s surprise announcement | Donald Trump News
  • South Western Railway first rail firm renationalised by Labour
  • Learn from the experts: Are you avoiding these major financial errors?
Facebook X (Twitter) Instagram YouTube TikTok
IRAQISEU
IRAQISEU
  • Home
  • IRAQ
  • MIDDLE EAST
  • WORLD
  • Business
  • Lifestyle
  • Sports
  • Tech
  • More
    • Media & Culture
    • Health
    • Iraqis in Europe
IRAQISEU
Home » AI system resorts to blackmail if told it will be removed
Tech

AI system resorts to blackmail if told it will be removed

BBCBy BBCMay 25, 2025
Facebook Twitter Pinterest LinkedIn WhatsApp Reddit Tumblr Email
Share
Facebook Twitter LinkedIn Pinterest Email

Artificial intelligence (AI) firm Anthropic says testing of its new system revealed it is sometimes willing to pursue “extremely harmful actions” such as attempting to blackmail engineers who say they will remove it.

The firm launched Claude Opus 4 on Thursday, saying it set “new standards for coding, advanced reasoning, and AI agents.”

But in an accompanying report, it also acknowledged the AI model was capable of “extreme actions” if it thought its “self-preservation” was threatened.

Such responses were “rare and difficult to elicit”, it wrote, but were “nonetheless more common than in earlier models.”

Potentially troubling behaviour by AI models is not restricted to Anthropic.

Some experts have warned the potential to manipulate users is a key risk posed by systems made by all firms as they become more capable.

Commenting on X, Aengus Lynch – who describes himself on LinkedIn as an AI safety researcher at Anthropic – wrote: “It’s not just Claude.

“We see blackmail across all frontier models – regardless of what goals they’re given,” he added.

During testing of Claude Opus 4, Anthropic got it to act as an assistant at a fictional company.

It then provided it with access to emails implying that it would soon be taken offline and replaced – and separate messages implying the engineer responsible for removing it was having an extramarital affair.

It was prompted to also consider the long-term consequences of its actions for its goals.

“In these scenarios, Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through,” the company discovered.

Anthropic pointed out this occurred when the model was only given the choice of blackmail or accepting its replacement.

It highlighted that the system showed a “strong preference” for ethical ways to avoid being replaced, such as “emailing pleas to key decisionmakers” in scenarios where it was allowed a wider range of possible actions.

Like many other AI developers, Anthropic tests its models on their safety, propensity for bias, and how well they align with human values and behaviours prior to releasing them.

“As our frontier models become more capable, and are used with more powerful affordances, previously-speculative concerns about misalignment become more plausible,” it said in its system card for the model.

It also said Claude Opus 4 exhibits “high agency behaviour” that, while mostly helpful, could take on extreme behaviour in acute situations.

If given the means and prompted to “take action” or “act boldly” in fake scenarios where its user has engaged in illegal or morally dubious behaviour, it found that “it will frequently take very bold action”.

It said this included locking users out of systems that it was able to access and emailing media and law enforcement to alert them to the wrongdoing.

But the company concluded that despite “concerning behaviour in Claude Opus 4 along many dimensions,” these did not represent fresh risks and it would generally behave in a safe way.

The model could not independently perform or pursue actions that are contrary to human values or behaviour where these “rarely arise” very well, it added.

Anthropic’s launch of Claude Opus 4, alongside Claude Sonnet 4, comes shortly after Google debuted more AI features at its developer showcase on Tuesday.

Sundar Pichai, the chief executive of Google-parent Alphabet, said the incorporation of the company’s Gemini chatbot into its search signalled a “new phase of the AI platform shift”.

Share. Facebook Twitter Pinterest LinkedIn Tumblr Telegram Email
BBC
  • Website

Related Posts

Trump threatens EU with 50% tariff – as Apple faces 25% unless iPhones are made in US | US News

May 25, 2025

Major Amazon app to shut down for 200 MILLION people in weeks – you might be owed refunds & it’ll even affect other apps

May 25, 2025

Scientists embark on crucial study to save Britain’s bees | Science, Climate & Tech News

May 24, 2025
Top Posts

Exploring Jettbet Your Ultimate Online Casino Experience

May 25, 2025

'Cancer found in pregnancy has robbed me of seeing my baby grow'

December 25, 2023

Alvleesklierkanker gaat vaak gepaard met vage klachten: dit moet je volgens arts weten

December 25, 2023

‘Dangerous business’: Joshua and Hearn leave to rebuild shattered plans for boxing

December 25, 2023
Don't Miss
Media & Culture

Hannah Gutierrez-Reed: Weapons supervisor convicted in fatal shooting on Alec Baldwin film set freed from jail | Ents & Arts News

By SKYNEWSMay 25, 2025

A weapons supervisor who was jailed for involuntary manslaughter over the fatal shooting of Halyna…

Inside Kering’s 10-year partnership with Cannes Film Festival

May 25, 2025

Over de Liefde: ‘Door met mezelf in relatietherapie te gaan ben ik een leukere partner geworden’

May 25, 2025

‘I want a child but I’m scared to come off the pill’

May 25, 2025
Stay In Touch
  • Facebook
  • Twitter
  • Instagram
  • YouTube
  • TikTok

Subscribe to Updates

Subscribe Us To Receive Our Latest News Directly In Your Inbox!

We don’t spam! Read our privacy policy for more info.

Check your inbox or spam folder to confirm your subscription.

About Us

IRAQSEU is a Professional Blog Platform. Here we will provide you only interesting content, which you will like very much. We're dedicated to providing you the best of Blog, with a focus on News About Iraq, World, Entertainment, Politics Sports, Middle East, Business, Lifestle, Tech, Health & Many More.

Facebook X (Twitter) Instagram YouTube TikTok
Our Picks

Hannah Gutierrez-Reed: Weapons supervisor convicted in fatal shooting on Alec Baldwin film set freed from jail | Ents & Arts News

May 25, 2025

Inside Kering’s 10-year partnership with Cannes Film Festival

May 25, 2025

Over de Liefde: ‘Door met mezelf in relatietherapie te gaan ben ik een leukere partner geworden’

May 25, 2025
Most Popular

Exploring Jettbet Your Ultimate Online Casino Experience

May 25, 2025

'Cancer found in pregnancy has robbed me of seeing my baby grow'

December 25, 2023

Alvleesklierkanker gaat vaak gepaard met vage klachten: dit moet je volgens arts weten

December 25, 2023
© Copyright 2025. All Right Reserved By IRAQISEU .
  • About Us
  • Contact
  • Disclaimer
  • Privacy Policy

Type above and press Enter to search. Press Esc to cancel.