TECH NEWS

PromptFiction Flaw Auto-Submitted Hidden Prompts in Claude Desktop

A single click on a crafted link could cause Claude Desktop to execute attacker-written instructions without prompting the user to review or send the prompt, according to new research from Oasis Security.

Securityaudit services

The vulnerability, named PromptFiction, affected Anthropic’s Claude Desktop application and used its custom claude:// URL scheme.

For context, a malicious link could open the app, submit a prepared prompt automatically, and instruct the AI agent to access sensitive information or perform actions through connected tools.

Anthropic fixed the flaw after it was reported through the company’s Responsible Disclosure Program.

However, before the correction, clicking a crafted link was enough to submit the prompt, with no confirmation screen or opportunity to inspect the instructions before Claude processed them.

Discover more

Machine Learning & Artificial Intelligence

Network security solutions

Engineering & Technology

A desktop link became an automatic command

Claude Desktop registers itself as a handler...

Copyright of this story solely belongs to hackread.com. To see the full text click HERE

https://platform.theverge.com/wp-content/uploads/sites/2/chorus/uploads/chorus_asset/file/25362061/STK_414_AI_CHATBOT_R2_CVirginia_D.jpg?quality=90&strip=all&crop=0%2C10.732984293194%2C100...

AI leaders sign statement asking the government to do something about automated AI

Employees of OpenAI and Anthropic, as well as Google, Meta, Thinking Machines, Microsoft, Mistral, and other leading AI labs, have written a statement to the US government supporting a potential slowdown of sorts for frontier AI development — or at least a speed-up of global coordinated governance efforts. “Al could help

https://image.theregister.com/5279934.jpg?imageId=5279934&x=0&y=0&cropw=100&croph=100&panox=0&panoy=0&panow=100&panoh=100&width=1200&height=683

Microsoft and Wiz mind-meld agents catch more than 90% of bugs

Secret to their success: Using the right model for the right security job Two agentic bug-hunting systems from Microsoft and Google-owned Wiz show that when it comes to finding and remediating software vulnerabilities, at least two models’ minds work better than one - and Wiz tells us it’s adding

https://image.theregister.com/259916.jpg?imageId=259916&x=0&y=0&cropw=100&croph=100&panox=0&panoy=0&panow=100&panoh=100&width=1200&height=683

War machines can run amok with AI in control

AIs not only make life-and-death targeting decisions on the battlefield, but also involve humans in only some stages of the decision-making process. To illustrate how that happens and prompt greater scrutiny of model-enabled killing, Airwars, a not-for-profit transparency watchdog, has published a detailed report, taking apart the ways machine learning

https://www.cnet.com/wp-content/uploads/sites/2/7f374ad2-93b1-42e3-90d3-e3607c40beb5.jpg

Best Mattress in a Box for 2026, Tested by Our Sleep Experts

Is it time to ditch that old, worn-out mattress? If you’re not getting quality sleep at night or are waing up with aches and pains, a new bed may make all of the difference. Upgrading your mattress no longer requires spending hours wandering around a store looking for the

A desktop link became an automatic command

Read more

AI leaders sign statement asking the government to do something about automated AI

Microsoft and Wiz mind-meld agents catch more than 90% of bugs

War machines can run amok with AI in control

Best Mattress in a Box for 2026, Tested by Our Sleep Experts