Posts of last few hours

In this interview with Help Net Security, Jonathan Care, Lead Analyst and Practice Lead AI at KuppingerCole Analysts, compares AI security startups to the early fraud market he helped define at Gartner. He explains which categories will survive as products and which will become features inside bigger platforms. Care talks about Q3 funding going to fewer, larger rounds, Israel’s share of cyber venture investment, and the five controls an organization needs before an agent touches … More →

The post Several AI security market categories are headed for bigger platforms appeared first on Help Net Security.

If you use Microsoft Authenticator on Android, update it. Researchers at Enclave found a flaw that let any app on your phone quietly take over your work account. The app didn’t need permissions, and you didn’t need to tap anything. Here’s what happened Authenticator does more than show six-digit codes. On Android it also acts as a “broker.” When Teams, Outlook or another Microsoft app needs to sign you in, it asks Authenticator, and Authenticator … More →

The post Microsoft Authenticator vulnerability let rogue Android apps steal work account access appeared first on Help Net Security.

Madhu Shashanka, Chief Scientist at Concentric AI, explains why securing LLM tools and agents depends more on your data than on the model. Controls on model behavior, such as guardrails and injection detection, lower the rate of bad outcomes but never reach zero. Controls on what goes in are different, because you own the documents, systems, and credentials an agent can reach. He covers prompt injection, data leakage through connectors, overprivileged agents, and shadow AI. … More →

The post LLM agent security needs more than guardrails appeared first on Help Net Security.

Amutable opened signups for the private preview of Amutable Foundations, a Linux system in which each machine sends a signed report of the software and configuration it runs, drawing on local hardware to make the report easier to verify. The company says its engineers write and maintain code in the kernel, in systemd and in image-building tools. If you run a Linux fleet, this changes what you check. Say a machine can’t show it matches … More →

The post Amutable Foundations is a Linux system with no SSH and no package manager appeared first on Help Net Security.

Attackers are exploiting newly disclosed security flaws within days, leaving organizations little time to respond, according to Rapid7’s Q3 2026 Quarterly Threat Landscape Report. The company reported an average of three days between public disclosure and exploitation of high-severity vulnerabilities, with 71% of tracked exploitation occurring within the first week. Cumulative exploitation rate (Source: Rapid7) Exploitation puts pressure on fixed patching schedules Researchers tracked 43 newly exploited high- and critical-severity vulnerabilities during the quarter. Organizations … More →

The post 71% of vulnerability exploitation starts within a week of disclosure appeared first on Help Net Security.

Cybersecurity Expert / Penetration Tester Barrion | Sweden | On-site – No longer accepting applications As a Cybersecurity Expert / Penetration Tester, you will test web applications and APIs for authentication, authorisation, and business logic vulnerabilities. You’ll identify complex attack paths and validate findings.You’ll evaluate AI-driven security testing, investigate false positives and missed vulnerabilities, and develop scenarios that improve automated discovery. Cloud Network Solution Architect TAL | Australia | On-site – View job details As … More →

The post Cybersecurity jobs available right now: October 27, 2026 appeared first on Help Net Security.

OpenAI released GPT-6.1 Terra, a model that operates desktop software, to all Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex, and to developers through its API as gpt-6.1-terra. API prices are 50 percent below those of GPT-5.6 Terra and GPT-6.1 Sol: $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens. Tokens are the chunks of text a model reads and writes. OpenAI says Terra … More →

The post GPT-6.1 Terra meets OpenAI’s Critical cyber threshold, and it’s cheaper than GPT-5.6 Terra appeared first on Help Net Security.