AI Sec Watch

Anthropic, an AI company, is starting its own drug discovery program to develop AI tools for pharmaceutical companies, focusing on treatments for neglected diseases (illnesses that traditional drug makers ignore because they're not profitable). The company believes that by doing its own drug research alongside pharmaceutical partners, it can build better AI products and get direct feedback on what drugmakers actually need.

Claude Sonnet 5 is a new AI model released by Anthropic with performance similar to their more advanced Opus 4.8 model but at lower advertised prices. However, the model uses a new tokenizer (a system that breaks text into small units for processing) that produces approximately 30% more tokens for the same input text, effectively increasing costs by 1.27x to 1.42x depending on the language, with minimal impact on Simplified Mandarin text.

A missing authorization vulnerability in GitHub Enterprise Server allowed authenticated users to read source code from private repositories they shouldn't have access to. The vulnerability existed in a Copilot pull request description feature that compared code across repositories without checking if the user had permission to view the target repository, and it required the attacker to already have read access to at least one repository on the system.

Cybersecurity companies Palo Alto Networks and CrowdStrike have experienced record growth due to increased demand for AI-powered security defenses, driven by concerns about Mythos-class models (advanced AI systems that hackers could use to find software vulnerabilities and launch attacks). Both companies have positioned themselves as leaders by gaining early access to powerful AI tools and acquiring identity security technology to help protect businesses from AI-driven cyber threats.

IBM Langflow OSS versions 1.0.0 through 1.10.0 have a security flaw where stored credentials can be exposed because the software uses weak encryption at rest (encryption that protects data when it's not being actively used). The problem stems from using a poor key derivation mechanism, which is the process that converts a password into the actual encryption key.

IBM Langflow OSS (an open-source AI framework) versions 1.0.0 through 1.10.0 has a vulnerability that lets authenticated users (those with login access) run arbitrary OS commands (any instructions on the computer's operating system) and access sensitive files like credentials, potentially compromising the entire system and allowing attackers to move to other connected systems. This is classified as a code injection flaw (CVE-2026-7873), where attackers can trick the application into executing malicious code.

IBM Langflow OSS (open-source software) versions 1.0.0 through 1.10.0 have a security flaw where users who can access Redis (a data storage system) can execute arbitrary code (run commands they shouldn't be able to run) with full control over the application, potentially exposing all stored secrets, data, and system integrity.

IBM Langflow OSS (an open-source tool for building AI workflows) versions 1.0.0 through 1.10.0 has a security flaw where it doesn't properly check the data in flow nodes (building blocks of a workflow) when their component type fields are missing or empty, which could allow an attacker to run arbitrary code (any commands they want) on the system. The vulnerability is caused by improper input validation (failing to check whether incoming data is safe and correct before using it).

IBM Langflow OSS (open-source software) versions 1.0.0 through 1.9.6 has a security flaw where attackers without login credentials can access protected project resources and perform operations through the Streamable MCP (model context protocol, a system for AI tools to communicate) transport endpoint because the software fails to properly check user permissions. This vulnerability affects the authorization layer, meaning the system doesn't adequately verify who is allowed to do what.