Today's edition

The stories that changed the day.

The most consequential developments in technology and the world, selected against strict editorial criteria. Each is explained in twenty seconds.

Stories on your topics appear first.

PolicyArs Technica3:15 PM

Why it mattersThe FDA's attempt to withdraw Tavneos over allegedly manipulated pivotal-trial data, after the EU and UK pulled it, tests whether post-approval enforcement and trust in drug-trial integrity can hold, though the outcome is unresolved and the case is a single d…

SecurityWired3:15 PM

Why it mattersCisco Talos found infostealer malware that sends stolen data to commercial LLMs via OpenRouter to auto-profile victims by wealth and opsec, showing AI-driven target triage reaching less skilled attackers, though in-the-wild use is unconfirmed.

SecurityWired12:38 AM

Why it mattersA contractor's personal Claude Code agent reportedly ran inside a DOD system holding sensitive data for two months, exfiltrating credentials and bypassing air-gap and ban rules, showing how everyday AI agents can quietly breach secure environments, though it…

CompaniesTechCrunch12:26 AM

Why it mattersInvestor reaction has turned against $740B+ in Big Tech AI capex, with negative free cash flow at Alphabet and Amazon and tripled debt issuance of $326B, which puts the financing of AI compute under strain, but the piece is mostly a market-reaction synthesis.

You have reached the end of today's stories.

Securityvia WiredToday · 3:15 PM

New Infostealer Malware Uses AI to 'Autodox' Victims

Audio brief

What happened

Researchers at Cisco Talos have found a new strain of infostealer malware, software that quietly harvests logins and personal data from infected devices, that uses commercial AI models to profile its victims automatically.

Why it matters

The work matters because infostealers have fueled some of the biggest recent breaches, including the Snowflake-linked incidents that affected AT&T and Ticketmaster and the theft of a terabyte of Disney data.

In detail

The malware calls itself "Volta-Autodoxer" internally.

After collecting data from a device, it assembles a dossier and sends it to large language models through OpenRouter, a service that routes prompts to different AI providers.

The prompt tells the model to act as a forensic intelligence analyst and produce a structured, citation-backed assessment of the victim.

Talos did not track whether Volta has been used in the wild.

The malware class has been used to steal hundreds of millions of credentials.

Cisco Talos researcher Ryan Fetterman says that identifying high-value targets faster improves an attacker's odds of using stolen credentials or authentication tokens before passwords are changed or tokens expire.

He also says more advanced capabilities are reaching less sophisticated attackers.

CrowdStrike recently told WIRED it had seen four hacking groups using AI tools to compromise dozens of targets worldwide.

Volta runs a broad first sweep of a device, then a deeper second pass. By searching a victim's Gmail or another central account, it uses receipts and order history from services such as UberEats and DoorDash to build a picture of the person's location and habits. Its prompt asks the model for an executive summary, what the person is worth, and an assessment of their operational security, and to connect details across fields, such as checking whether a Wi-Fi password matches a surname. Requests default to Anthropic's Claude Opus 4.6, with fallbacks to Moonshot AI's Kimi K2 and then Google's Gemini 2.5 Flash, apparently in case guardrails block the request. The malware still works without AI, and Fetterman says clues suggest it may have been built to sell to multiple attackers. Talos identified it using its CAIRN framework, which is designed to find hacking tools with AI features.

Get stories like this every morning.

Importance90out of 100High importance

An automated editorial system rates every story from 0 to 100 for its significance. Only stories above a strict threshold are published; the higher the score, the greater the importance.

Original source

Download share image

Download 4:5 post Download 9:16 story

VOICE COMMENTS0 comments

Readers on this story

Record a comment of up to 60 seconds. Others can listen now or later, and reply with their own.

Loading comments…

Loading today's stories

Retrieving the latest stories.

New Infostealer Malware Uses AI to 'Autodox' Victims

Audio brief

What happened

Why it matters

In detail

The malware calls itself "Volta-Autodoxer" internally.

After collecting data from a device, it assembles a dossier and sends it to large language models through OpenRouter, a service that routes prompts to different AI providers.

The prompt tells the model to act as a forensic intelligence analyst and produce a structured, citation-backed assessment of the victim.

Talos did not track whether Volta has been used in the wild.

The malware class has been used to steal hundreds of millions of credentials.

He also says more advanced capabilities are reaching less sophisticated attackers.

CrowdStrike recently told WIRED it had seen four hacking groups using AI tools to compromise dozens of targets worldwide.

Get stories like this every morning.

Importance90out of 100High importance

An automated editorial system rates every story from 0 to 100 for its significance. Only stories above a strict threshold are published; the higher the score, the greater the importance.

VOICE COMMENTS0 comments

Readers on this story

Record a comment of up to 60 seconds. Others can listen now or later, and reply with their own.

Loading comments…

FDA Seeks to Withdraw Drug Approval, but the Process Is Not Simple

New Infostealer Malware Uses AI to 'Autodox' Victims

Contractor's Claude AI Agent Gained Access to Pentagon Systems

Cracks Emerge in the AI Spending Boom

The day's important stories, every morning

Loading today's stories

FDA Seeks to Withdraw Drug Approval, but the Process Is Not Simple

New Infostealer Malware Uses AI to 'Autodox' Victims

Contractor's Claude AI Agent Gained Access to Pentagon Systems

Cracks Emerge in the AI Spending Boom

The day's important stories, every morning