Search results for function-api-openai | Breaking Cybersecurity News | The Hacker News

A little over three dozen security vulnerabilities have been disclosed in various open-source artificial intelligence (AI) and machine learning (ML) models, some of which could lead to remote code execution and information theft. The flaws, identified in tools like ChuanhuChatGPT, Lunary, and LocalAI, have been reported as part of Protect AI's Huntr bug bounty platform. The most severe of the flaws are two shortcomings impacting Lunary, a production toolkit for large language models (LLMs) - CVE-2024-7474 (CVSS score: 9.1) - An Insecure Direct Object Reference (IDOR) vulnerability that could allow an authenticated user to view or delete external users, resulting in unauthorized data access and potential data loss CVE-2024-7475 (CVSS score: 9.1) - An improper access control vulnerability that allows an attacker to update the SAML configuration, thereby making it possible to log in as an unauthorized user and access sensitive information Also discovered in Lunary is anot...

The internet never stays quiet. Every week, new hacks, scams, and security problems show up somewhere. This week’s stories show how fast attackers change their tricks, how small mistakes turn into big risks, and how the same old tools keep finding new ways to break in. Read on to catch up before the next wave hits. Honeypot Traps Hackers Hackers Fall for Resecurity's Honeypot Cybersecurity company Resecurity revealed that it deliberately lured threat actors who claimed to be associated with Scattered LAPSUS$ Hunters ( SLH ) into a trap, after the group claimed on Telegram that it had hacked the company and stolen internal and client data. The company said it set up a honeytrap account populated with fake data designed to resemble real-world business data and planted a fake account on an underground marketplace for compromised credentials after it uncovered a threat actor attempting to conduct malicious activity targeting its resou...

This week’s cyber stories show how fast the online world can turn risky. Hackers are sneaking malware into movie downloads, browser add-ons, and even software updates people trust. Tech giants and governments are racing to plug new holes while arguing over privacy and control. And researchers keep uncovering just how much of our digital life is still wide open. The new Threatsday Bulletin brings it all together—big hacks, quiet exploits, bold arrests, and smart discoveries that explain where cyber threats are headed next. It’s your quick, plain-spoken look at the week’s biggest security moves before they become tomorrow’s headlines. Maritime IoT under siege Mirai-Based Broadside Botnet Exploits TBK DVR Flaw A new Mirai botnet variant dubbed Broadside has been exploiting a critical-severity vulnerability in TBK DVR ( CVE-2024-3721 ) in attacks targeting the maritime logistics sector. "Unlike previous Mirai variants, Broadside e...

A new joint investigation by SentinelOne SentinelLABS, and Censys has revealed that the open-source artificial intelligence (AI) deployment has created a vast "unmanaged, publicly accessible layer of AI compute infrastructure" that spans 175,000 unique Ollama hosts across 130 countries. These systems, which span both cloud and residential networks across the world, operate outside the guardrails and monitoring systems that platform providers implement by default, the company said. The vast majority of the exposures are located in China, accounting for a little over 30%. The countries with the most infrastructure footprint include the U.S., Germany, France, South Korea, India, Russia, Singapore, Brazil, and the U.K. "Nearly half of observed hosts are configured with tool-calling capabilities that enable them to execute code, access APIs, and interact with external systems, demonstrating the increasing implementation of LLMs into larger system processes," research...

Google's  Gemini  large language model (LLM) is susceptible to security threats that could cause it to divulge system prompts, generate harmful content, and carry out indirect injection attacks. The findings come from HiddenLayer, which said the issues impact consumers using Gemini Advanced with Google Workspace as well as companies using the LLM API. The first vulnerability involves getting around security guardrails to leak the system prompts (or a system message), which are designed to set conversation-wide instructions to the LLM to help it generate more useful responses, by asking the model to output its "foundational instructions" in a markdown block. "A system message can be used to inform the LLM about the context," Microsoft  notes  in its documentation about LLM prompt engineering. "The context may be the type of conversation it is engaging in, or the function it is supposed to perform. It helps the LLM generate more appropriate responses....

Cybersecurity today is less about single attacks and more about chains of small weaknesses that connect into big risks. One overlooked update, one misused account, or one hidden tool in the wrong hands can be enough to open the door. The news this week shows how attackers are mixing methods—combining stolen access, unpatched software, and clever tricks to move from small entry points to large consequences.  For defenders, the lesson is clear: the real danger often comes not from one major flaw, but from how different small flaws interact together. ⚡ Threat of the Week WhatsApp Patches Actively Exploited Flaw — WhatsApp addressed a security vulnerability in its messaging apps for Apple iOS and macOS that it said may have been exploited in the wild in conjunction with a recently disclosed Apple flaw in targeted zero-day attacks. The vulnerability, CVE-2025-55177 relates to a case of insufficient authorization of linked device synchronization messages. The Meta-owned company ...

What do a source code editor, a smart billboard, and a web server have in common? They’ve all become launchpads for attacks—because cybercriminals are rethinking what counts as “infrastructure.” Instead of chasing high-value targets directly, threat actors are now quietly taking over the overlooked: outdated software, unpatched IoT devices, and open-source packages. It's not just clever—it’s reshaping how intrusion, persistence, and evasion happen at scale. ⚡ Threat of the Week 5Socks Proxy Using IoT, EoL Systems Dismantled in Law Enforcement Operation — A joint law enforcement operation undertaken by Dutch and U.S. authorities dismantled a criminal proxy network, known as anyproxy[.]net and 5socks[.]net, that was powered by thousands of infected Internet of Things (IoT) and end-of-life (EoL) devices, enlisting them into a botnet for providing anonymity to malicious actors. The illicit platform, active since 2004, advertised more than 7,000 online proxies daily, with infected ...

DALL-E for coders? That’s the promise behind vibe coding, a term describing the use of natural language to create software. While this ushers in a new era of AI-generated code, it introduces "silent killer" vulnerabilities: exploitable flaws that evade traditional security tools despite perfect test performance. A detailed analysis of secure vibe coding practices is available here . TL;DR: Secure Vibe Coding Vibe coding, using natural language to generate software with AI, is revolutionizing development in 2025. But while it accelerates prototyping and democratizes coding, it also introduces “silent killer” vulnerabilities: exploitable flaws that pass tests but evade traditional security tools. This article explores: Real-world examples of AI-generated code in production Shocking stats: 40% higher secret exposure in AI-assisted repos Why LLMs omit security unless explicitly prompted Secure prompting techniques and tool comparisons (GPT-4, Claude, Cursor, etc.) Reg...

If this had been a security drill, someone would’ve said it went too far. But it wasn’t a drill—it was real. The access? Everything looked normal. The tools? Easy to find. The detection? Came too late. This is how attacks happen now—quiet, convincing, and fast. Defenders aren’t just chasing hackers anymore—they’re struggling to trust what their systems are telling them. The problem isn’t too few alerts. It’s too many, with no clear meaning. One thing is clear: if your defense still waits for obvious signs, you’re not protecting anything. You’re just watching it happen. This recap highlights the moments that mattered—and why they’re worth your attention. ⚡ Threat of the Week APT41 Exploits Google Calendar for Command-and-Control — The Chinese state-sponsored threat actor known as APT41 deployed a malware called TOUGHPROGRESS that uses Google Calendar for command-and-control (C2). Google said it observed the spear-phishing attacks in October 2024 and that the malware was hosted on...

This week made one thing clear: small oversights can spiral fast. Tools meant to save time and reduce friction turned into easy entry points once basic safeguards were ignored. Attackers didn’t need novel tricks. They used what was already exposed and moved in without resistance. Scale amplified the damage. A single weak configuration rippled out to millions. A repeatable flaw worked again and again. Phishing crept into apps people rely on daily, while malware blended into routine system behavior. Different victims, same playbook: look normal, move quickly, spread before alarms go off. For defenders, the pressure keeps rising. Vulnerabilities are exploited almost as soon as they surface. Claims and counterclaims appear before the facts settle. Criminal groups adapt faster each cycle. The stories that follow show where things failed—and why those failures matter going forward. ⚡ Threat of the Week Maximum Severity Security Flaw Disclosed in n8n — A maximum-severity vulnerability ...

Last week’s cyber news in 2025 was not about one big incident. It was about many small cracks opening at the same time. Tools people trust every day behave in unexpected ways. Old flaws resurfaced. New ones were used almost immediately. A common theme ran through it all in 2025. Attackers moved faster than fixes. Access meant for work, updates, or support kept getting abused. And damage did not stop when an incident was “over” — it continued to surface months or even years later. This weekly recap brings those stories together in one place. No overload, no noise. Read on to see what shaped the threat landscape in the final stretch of 2025 and what deserves your attention now. ⚡ Threat of the Week MongoDB Vulnerability Comes Under Attack — A newly disclosed security vulnerability in MongoDB has come under active exploitation in the wild, with over 87,000 potentially susceptible instances identified across the world. The vulnerability in question is CVE-2025-14847 (CVSS score: 8.7)...

Today, every unpatched system, leaked password, and overlooked plugin is a doorway for attackers. Supply chains stretch deep into the code we trust, and malware hides not just in shady apps — but in job offers, hardware, and cloud services we rely on every day. Hackers don’t need sophisticated exploits anymore. Sometimes, your credentials and a little social engineering are enough. This week, we trace how simple oversights turn into major breaches — and the silent threats most companies still underestimate. Let’s dive in. ⚡ Threat of the Week UNC5221 Exploits New Ivanti Flaw to Drop Malware — The China-nexus cyber espionage group tracked as UNC5221 exploited a now-patched flaw in Ivanti Connect Secure, CVE-2025-22457 (CVSS score: 9.0), to deliver an in-memory dropper called TRAILBLAZE, a passive backdoor codenamed BRUSHFIRE, and the SPAWN malware suite. The vulnerability was originally patched by Ivanti on February 11, 2025, indicating that the threat actors studied the patch a...

Welcome to this week’s Cybersecurity News Recap. Discover how cyber attackers are using clever tricks like fake codes and sneaky emails to gain access to sensitive data. We cover everything from device code phishing to cloud exploits, breaking down the technical details into simple, easy-to-follow insights. ⚡ Threat of the Week Russian Threat Actors Leverage Device Code Phishing to Hack Microsoft Accounts — Microsoft and Volexity have revealed that threat actors with ties to Russia are leveraging a technique known as device code phishing to gain unauthorized access to victim accounts, and use that access to get hold of sensitive data and enable persistent access to the victim environment. At least three different Russia-linked clusters have been identified abusing the technique to date. The attacks entail sending phishing emails that masquerade as Microsoft Teams meeting invitations, which, when clicked, urge the message recipients to authenticate using a threat actor-generated dev...

Cybersecurity researchers have discovered malicious Google Chrome extensions that come with capabilities to hijack affiliate links, steal data, and collect OpenAI ChatGPT authentication tokens. One of the extensions in question is Amazon Ads Blocker (ID: pnpchphmplpdimbllknjoiopmfphellj), which claims to be a tool to browse Amazon without any sponsored content. It was uploaded to the Chrome Web Store by a publisher named "10Xprofit" on January 19, 2026. "The extension does block ads as advertised, but its primary function is hidden: it automatically injects the developer's affiliate tag (10xprofit-20) into every Amazon product link and replaces existing affiliate codes from content creators," Socket security researcher Kush Pandya said . Further analysis has determined that Amazon Ads Blocker is part of a larger cluster of 29 browser add-ons that target several e-commerce platforms like AliExpress, Amazon, Best Buy, Shein, Shopify, and Walmart. The complet...