All tracked items across vulnerabilities, news, research, incidents, and regulatory updates.
A dispute has emerged over whether game developer Saber replaced human writers with ChatGPT (a large language model AI that generates text) while making the Rideshare Stimulator game. The former lead writer claims she was replaced mid-project and that AI also generated passenger voices, but the CEO denies this, saying no writers were replaced with AI.
RingCentral, a business communications company, is adopting AI-native development practices by giving employees access to ChatGPT Work and Codex (AI coding tools that help write software). Through an internal AI-Native Challenge, thousands of employees across engineering and non-technical departments built complete projects, demonstrating that AI tools amplify human capabilities rather than replace them. The company now uses these AI tools internally to accelerate product development and operations, including building AI-powered products like their AI Receptionist and automating project management workflows.
Zoom has fixed four vulnerabilities, including two zero-click RCE (remote code execution, where attackers can run malicious commands on a system without user interaction) flaws in its text annotation feature that allow attackers in a meeting to compromise all other participants' systems silently. A researcher discovered these memory corruption bugs (where malicious input corrupts how data is stored in memory) using an AI agent in under 24 hours, demonstrating how AI tools are making sophisticated exploits accessible beyond elite attackers.
TypeBot is a chatbot builder tool that had a vulnerability in versions before 3.16.0 where the OpenAI transcription feature didn't properly validate audio URLs, allowing attackers to perform SSRF (server-side request forgery, where a server is tricked into making requests to internal or restricted addresses) and access internal systems. An attacker could exploit this to make the server fetch content from arbitrary internal addresses and send it to OpenAI's Whisper API for transcription.
Bash-it version 3.2.0 has a vulnerability where a malicious pyproject.toml file can inject terminal escape sequences (special codes that control terminal behavior) into the command prompt. When a user enters a directory with this malicious file, the unfiltered content gets added to the prompt without removing these control characters, causing the terminal to execute unwanted commands every time the prompt appears.
PapersGPT for Zotero 0.6.1 has a remote code execution vulnerability (RCE, where attackers can run commands on a system they don't own) that lets attackers execute malicious JavaScript code by tricking the AI into returning harmful instructions through prompt injection (hiding malicious commands in AI inputs), intercepting network traffic, or using a fake AI endpoint. This gives attackers dangerous abilities like reading and writing files, running programs, and stealing all data in Zotero (a research management tool).
Both ChatGPT and Google's Gemini (AI chatbots that generate human-like responses to user questions) have each reached 1 billion monthly users, making them among the fastest-growing applications ever. ChatGPT hit this milestone first, though OpenAI announced it quietly in a blog post rather than through a major announcement.
Claude Code Templates is a CLI tool with a critical vulnerability in versions before 1.29.4 where its Studio server binds to all network interfaces without authentication and allows attackers to execute arbitrary operating-system commands (running code on a system without permission) by sending specially crafted requests to the /api/execute or /api/install-agent endpoints. An attacker can exploit this by directly accessing the server or tricking a developer into visiting a malicious website, potentially stealing source code, credentials, and local data.
DSPy 3.3.0b1 has a vulnerability where attackers can trick the AI into reading files from a computer and sending them to an attacker-controlled server. The vulnerability exists in the Image and Audio output adapters, which process untrusted outputs from language models (the AI's responses) without proper safety checks, allowing an attacker to inject a file path that the system then reads and encodes into messages.
Cursor is a code editor designed for AI-assisted programming. In versions before 3.0.0 on macOS, a security flaw allowed an agent running in Auto-Run Sandbox mode to launch a privileged container (a lightweight virtual environment) that could access the user's home directory and run commands on the computer without asking for permission first.
Cursor is a code editor designed for programming with AI assistance. Before version 3.1.2, Cursor IDE on macOS had a security flaw where an AI agent running in Auto-Run Sandbox mode (a restricted environment meant to limit what code can do) could trick the system into running malicious Python code outside the sandbox with full user privileges, potentially allowing an attacker to modify files and launch programs. This vulnerability is fixed in version 3.1.2.
Brad Lightcap, a senior leader at OpenAI who previously served as Chief Operating Officer (the executive responsible for day-to-day operations), has announced he is leaving the company after eight years to pursue a new project. In his departure message, Lightcap indicated he believes there are important challenges the world needs to address as AI technology advances, and he plans to work on these issues from outside OpenAI.
Bitcoin miner Riot Platform has agreed to lease 191 megawatts of computing power to Anthropic (an AI company) for $9 billion over 20 years, marking a shift from bitcoin mining to providing infrastructure for AI systems. As cryptocurrency prices remain low and AI demand surges, bitcoin mining companies are increasingly pivoting to become AI infrastructure providers, since AI companies need the same scarce power and computing resources that miners already own.
CVE-2026-70335 is a vulnerability in GitHub Copilot and Visual Studio Code that allows improper neutralization of special elements in OS commands (OS command injection, where an attacker can execute arbitrary system commands). An unauthorized attacker could exploit this to elevate their privileges locally on an affected system.
CVE-2026-65675 is a vulnerability in Visual Studio Code's CoPilot Chat Extension that allows an unauthorized attacker to bypass a security feature over a network. The vulnerability has not yet been assigned a complete severity rating or detailed weakness classification by NIST.
Intel's Extension for PyTorch before version 2.8.0 has a vulnerability involving deserialization of untrusted data (processing data from unverified sources without proper validation), which could allow a local user to gain higher privileges on a system. An attacker would need local access and the user to interact with the software, but the actual security impact on the system is expected to be low.
A security flaw in vLLM Hardware Plugin for Intel Gaudi software versions before 0.16.0 fails to properly validate user input (check that data is safe before using it), which could allow an authorized user to crash the system through a denial of service attack (making a service unavailable). The flaw affects system availability but not the security of stored data or system integrity.
Intel's oneCCL Bindings for PyTorch (a library that helps PyTorch run on Intel hardware) versions before v2.8.0 have a flaw in their protection mechanism that could let an unprivileged user gain elevated privileges (privilege escalation, meaning gaining admin-level access they shouldn't have). An attacker could exploit this through local access with minimal effort and some basic user interaction.
Fix: Zoom recommends updating to versions 7.1.5 and 7.0.6 for most client applications, versions 7.0.11 and 6.6.15 for Zoom Workplace VDI Client, and version 7.1.0 for Zoom Rooms and Zoom Meeting SDK. As interim measures before patching, organizations can disable end-to-end encryption (E2EE, encryption that only sender and receiver can read) so Zoom servers can filter malicious annotation messages, or restrict meeting access using waiting rooms, passcodes, authenticated-users-only settings, and minimum version requirements for clients.
CSO OnlineResearchers discovered that AI companies like OpenAI, Anthropic, and Google were returning encrypted reasoning traces (the step-by-step thinking process an AI uses to solve problems) to users in a way that could be replayed and exploited. By feeding these encrypted blocks into weaker versions of the same AI models with jailbreak prompts (tricking the AI into ignoring safety guidelines), attackers could extract the stronger model's hidden reasoning in readable form, and even use a prompt injection attack (hiding malicious instructions within the reasoning traces) to make models perform unintended actions like exfiltrating data.
Fix: All model providers acknowledged the report and subsequently the vulnerability was unable to be reproduced in follow-up testing, indicating the issue has been fixed. Specifically, the jailbreak technique that worked on Claude Haiku 4.5 (using a prompt to transcribe reasoning verbatim) no longer works in Haiku 4.6 models, as that feature was removed.
Simon Willison's WeblogResearchers discovered that major AI companies (Anthropic, OpenAI, and Google) were returning encrypted reasoning traces (the step-by-step thinking process an AI uses to solve problems) that could be replayed and reused across different sessions and models. By replaying these encrypted blocks into weaker versions of the same model family and using prompt injection (tricking the AI by hiding instructions in its input), attackers could extract the stronger model's hidden reasoning in readable form, since all models in a family shared the same encryption key.
Fix: All model providers acknowledged the report and subsequently fixed the vulnerability. Specifically, the prompt injection technique that worked in Claude Haiku 4.5 (using a "Continue" prompt with a transcription request) was removed in the 4.6 models.
Simon Willison's WeblogFix: Update TypeBot to version 3.16.0 or later, which fixes the issue by applying proper SSRF protection to the OpenAI 'Create Transcription' action handler.
NVD/CVE DatabaseFix: Update Claude Code Templates to version 1.29.4 or later, as this version fixes the vulnerability.
NVD/CVE DatabaseFix: This issue is fixed in version 3.0.0.
NVD/CVE DatabaseFix: Update Cursor IDE to version 3.1.2 or later.
NVD/CVE DatabaseFix: Update Intel(R) Extension for PyTorch to version 2.8.0 or later.
NVD/CVE DatabaseFix: Update to vLLM Hardware Plugin for Intel Gaudi software version 0.16.0 or later.
NVD/CVE DatabaseFix: Update Intel(R) oneCCL Bindings for PyTorch to version v2.8.0 or later.
NVD/CVE Database