The Hacker News چهارشنبه ۲۰ آبان ۱۴۰۵ · ۱۵:۳۱

We Built An Agent That Broke All of Our Security Rules. Here's What We Learned.

We tell customers three things about securing AI agents: they need their own identities, access that fits the job they’re given, and traceability that’s separate from any human’s. That was true for us at Token as well. Then, we built an agent that broke all three by default. This helped us think through how such agents should be defended. When a browser agent acts inside a user’s authenticated

Read the full article at The Hacker News Removal request

This page shows only the headline and the short summary. The full text is published by the original source.

Related news

Full archive →

The Hacker News چهارشنبه ۲۰ آبان ۱۴۰۵ · ۱۵:۳۰

eMudhra Revokes 8 More CAs It Previously Declared Unaffected by Key Compromise

The in-house Python script behind eMudhra's October key compromise has now cost the company 8 more of its intermediate certificate authorities (CAs), even though it did not weaken their keys. Each intermediate CA has its own private key, which the company uses to sign the certificates it issues. The 8 intermediate CAs compromised in October had a different problem: a single line in the script

Programming & Languages

Source

The Hacker News چهارشنبه ۲۰ آبان ۱۴۰۵ · ۱۱:۴۷

Microsoft Patches 1,259 Flaws, Including Exploited Outlook Bug That May Expose NTLMv2 Hashes

Microsoft on Tuesday released patches for a total of 1,259 vulnerabilities spanning its software portfolio, the second consecutive month where the patch volume has exceeded 1,000 flaws. These cover 726 flaws in Windows, 174 in Office, 61 in Developer Tools, 59 in SQL Server, 53 in Developer Tools, and 27 in SharePoint Server. The patches also include 165 non-Microsoft-issued CVE identifiers.

Vulnerabilities & Zero-Day Databases Operating Systems

Source

The Hacker News چهارشنبه ۲۰ آبان ۱۴۰۵ · ۱۱:۱۹

Railway Says Stolen Customer AI Keys Led to $740,000 in Estimated Unauthorized Usage Charges

AI provider keys that customers stored on Railway were stolen in an October breach and used to commit an estimated $740,000 in unauthorized spending, the platform-as-a-service provider said yesterday. It has promised to pay those charges itself when an AI provider will not refund them. The company builds and runs apps for its customers and stores their secrets, such as API keys and database

Data Breach Databases

Source

The Hacker News چهارشنبه ۲۰ آبان ۱۴۰۵ · ۱۰:۵۱

Microsoft Researchers Show AI Agents Can Be Backdoored by Editing Fewer Than 10 Neurons

An AI model can be given a hidden backdoor by editing fewer than 10 of its rarely used neurons, without retraining the model, Microsoft researchers have shown. When a chosen trigger appears, the model outputs an attacker's command, which an agent using that model would run with the agent's own permissions. The attacker must be able to change the model's weights. The team tested the method on

Malware & Threats

Source