Author: HackWatchit

LLMs, Agents, Tools, and Frameworks Generative Artificial intelligence (GenAI) is full of technical concepts and terms; a few terms we often encounter are Large Language Models (LLMs), AI agents, and agentic systems. Although related, they serve different (but related) purposes within the AI ecosystem. LLMs are the foundational language engines designed to process and generate text (and images in the case of multi-model ones), while agents are meant to extend LLMs’ capabilities by incorporating tools and strategies to tackle complex problems effectively. Properly designed and built agents can adapt based on feedback, refining their plans and improving performance to try…

has a wide range of use cases in the natural language processing (NLP) domain, such as document tagging, survey analysis, and content organization. It falls under the realm of unsupervised learning technique, making it a very cost-effective technique that reduces the resources required to collect human-annotated data. We will dive deeper into BERTopic, a popular python library for transformer-based topic modeling, to help us process financial news faster and reveal how the trending topics change overtime.BERTopic consists of 6 core modules that can be customized to suit different use cases. In this article, we’ll examine, experiment with each module individually…

Jordan Drysdale // tl;dr Cisco Smart Install is awesome (on by default)…for hackers… not sysadmins. So, you Nessus too? Criticals and highs are all that matter! Right??? Until this beauty came along, but wait….this isn’t Critical or High. Let’s just ignore it. What can we do with this thing? Download config Upload config Execute commands How do we find it? nmap -p4786 0.0.0.0/0    ### this is bad! Probably don’t do this. So what? Let’s grab SIET (https://github.com/Sab0tag3d/SIET) and gather a config. ./siet.py -i 10.0.0.1 -g And yeah, you guessed it, we managed to download a config file. But, so…

LockBit’s dark web domains were hacked, exposing internal data, affiliate tools, and over 60,000 Bitcoin wallets in a major blow to the ransomware group. The ransomware group LockBit has suffered a serious blow after its dark web domain and infrastructure was hijacked and defaced by an unknown attacker. The breach, which surfaced on May 7, disrupted several of the group’s hidden service panels and included a public taunt: “Don’t do crime. CRIME IS BAD xoxo from Prague.” This breach is the latest in a series of setbacks for LockBit, once considered one of the most prolific ransomware-as-a-service (RaaS) operations. It…

In brief Kuwait has questioned 116 over illegal crypto mining as part of a crackdown. Authorities are investigating the illegal use of electricity for crypto mining across 59 residential properties. Kuwaiti residents pay just 4.6 cents per kilowatt-hour of electricity, four times less than the U.S. average. Kuwaiti authorities have ramped up their crackdown on illegal crypto mining, launching investigations into 31 cases and questioning 116.According to reports in local media, the investigations involve the illegal use of electricity for crypto mining across 59 residential properties, which authorities claim have strained the national power grid and caused widespread blackouts.Quoted in…

Ransomware operations are using legitimate Kickidler employee monitoring software for reconnaissance, tracking their victims’ activity, and harvesting credentials after breaching their networks. In attacks observed by cybersecurity companies Varonis and Synacktiv, Qilin and Hunters International ransomware affiliates installed Kickidler, an employee monitoring tool that can capture keystrokes, take screenshots, and create videos of the screen. Kickidler’s developer says the tool is used by over 5,000 organizations from 60 countries and provides visual monitoring and data loss prevention features. The attacks started with the threat actors taking out Google Ads displayed when people searched for RVTools, a free Windows utility for…

While the shortest distance between two points is a straight line, a straight-line attack on a large language model isn’t always the most efficient — and least noisy — way to get the LLM to do bad things. That’s why malicious actors have been turning to indirect prompt injection attacks on LLMs. Indirect prompt injection attacks involve malicious instructions embedded within external content — documents, web pages, or emails — that an LLM processes. The model may interpret these instructions as valid user commands, leading to unintended actions such as data leaks or misinformation. A team of researchers recently wrote…

Cisco is making significant strides in quantum computing by focusing on quantum networking, aiming to bring practical applications closer to reality. The company recently introduced a prototype of its Quantum Network Entanglement Chip and inaugurated the Cisco Quantum Lab in Santa Monica, California. The Quantum Network Entanglement Chip is designed to facilitate the connection of quantum processors, addressing the current limitation where most quantum processors have only hundreds of qubits, while practical applications require millions. “Even the most ambitious quantum computing roadmaps currently only target a few thousand qubits by 2030,” said Vijoy Pandey, SVP, Outshift by Cisco. By enabling…

Medical technology company Masimo Corporation disclosed that it experienced unauthorized activity on its on-premise network, affecting manufacturing operations. The cyber breach, identified on April 27, prompted the company to activate its incident response protocols, including isolating affected systems and initiating an investigation with third-party cybersecurity experts. As a result of the incident, some of Masimo’s manufacturing facilities have been operating below normal capacity, temporarily affecting the company’s ability to process, fulfill, and ship customer orders. The company is working diligently to restore normal operations and mitigate the impact of the breach. “We promptly commenced an investigation and are actively working…