AI agents can leak company data through simple web searches 2025-10-29 at 10:24 By Mirko Zorz When a company deploys an AI agent that can search the web and access internal documents, most teams assume the agent is simply working as intended. New research shows how that same setup can be used to quietly pull […]
React to this headline:
AI agents can leak company data through simple web searches Read More »
Scammers target international students by threatening their visa status 2025-10-29 at 08:29 By Sinisa Markovic In 2025, the U.S. government revoked thousands of visas from international students, often without warning or explanation. According to a newly released study, this opened a door for scammers. Posing as government officials, police, or university staff, they took advantage
React to this headline:
Scammers target international students by threatening their visa status Read More »
Can your earbuds recognize you? Researchers are working on it 2025-10-27 at 09:05 By Mirko Zorz Biometric authentication has moved from fingerprints to voices to facial scans, but a team of researchers believes the next step could be inside the ear. New research explores how the ear canal’s unique acoustic properties can be used to
React to this headline:
Can your earbuds recognize you? Researchers are working on it Read More »
Faster LLM tool routing comes with new security considerations 2025-10-23 at 09:23 By Sinisa Markovic Large language models depend on outside tools to perform real-world tasks, but connecting them to those tools often slows them down or causes failures. A new study from the University of Hong Kong proposes a way to fix that. The
React to this headline:
Faster LLM tool routing comes with new security considerations Read More »
AI’s split personality: Solving crimes while helping conceal them 2025-10-21 at 08:52 By Sinisa Markovic What happens when investigators and cybercriminals start using the same technology? AI is now doing both, helping law enforcement trace attacks while also being tested for its ability to conceal them. A new study from the University of Cagliari digs
React to this headline:
AI’s split personality: Solving crimes while helping conceal them Read More »
Most AI privacy research looks the wrong way 2025-10-20 at 13:19 By Mirko Zorz Most research on LLM privacy has focused on the wrong problem, according to a new paper by researchers from Carnegie Mellon University and Northeastern University. The authors argue that while most technical studies target data memorization, the biggest risks come from
React to this headline:
Most AI privacy research looks the wrong way Read More »
A new approach to blockchain spam: Local reputation over global rules 2025-10-17 at 10:18 By Mirko Zorz Spam has long been a nuisance in blockchain networks, clogging transaction queues and driving up fees. A new research paper from Delft University of Technology introduces a decentralized solution called STARVESPAM that could help nodes in permissionless blockchains
React to this headline:
A new approach to blockchain spam: Local reputation over global rules Read More »
When trusted AI connections turn hostile 2025-10-16 at 09:02 By Mirko Zorz Researchers have revealed a new security blind spot in how LLM applications connect to external systems. Their study shows that malicious Model Context Protocol (MCP) servers can quietly take control of hosts, manipulate LLM behavior, and deceive users, all while staying undetected by
React to this headline:
When trusted AI connections turn hostile Read More »
Humanoid robot found vulnerable to Bluetooth hack, data leaks to China 2025-10-16 at 07:33 By Sinisa Markovic Alias Robotics has published an analysis of the Unitree G1 humanoid robot, concluding that the device can be exploited as a tool for espionage and cyber attacks. A robot that can be hacked through Bluetooth Their tests show
React to this headline:
Humanoid robot found vulnerable to Bluetooth hack, data leaks to China Read More »
What if your privacy tools could learn as they go? 2025-10-14 at 11:49 By Mirko Zorz A new academic study proposes a way to design privacy mechanisms that can make use of prior knowledge about how data is distributed, even when that information is incomplete. The method allows privacy guarantees to stay mathematically sound while
React to this headline:
What if your privacy tools could learn as they go? Read More »
From theory to training: Lessons in making NICE usable 2025-10-10 at 09:02 By Mirko Zorz SMBs may not have big budgets, but they are on the receiving end of many cyberattacks. A new study from Cleveland State University looked at how these companies could train staff without getting lost in the thousands of skills and
React to this headline:
From theory to training: Lessons in making NICE usable Read More »
Rethinking AI security architectures beyond Earth 2025-10-08 at 09:39 By Mirko Zorz If you think managing cloud security is complex, try doing it across hundreds of satellites orbiting the planet. Each one is a moving endpoint that must stay secure while communicating through long, delay-prone links. A new study explores how AI could automate security
React to this headline:
Rethinking AI security architectures beyond Earth Read More »
New system aims to keep people connected when networks fail 2025-10-08 at 07:32 By Sinisa Markovic When disaster strikes, communication often fails. Cell towers can go offline, internet connections can disappear, and people are left without a way to share information or ask for help. A new research project looks at how to keep people
React to this headline:
New system aims to keep people connected when networks fail Read More »
Meet ARGUS, the robot built to catch hackers and physical intruders 2025-10-06 at 08:38 By Sinisa Markovic Hospitals, airports, and campuses are no longer dealing with separate security problems. Someone can slip past a checkpoint while another actor launches a network scan, and together those actions create a bigger risk than either one alone. Most
React to this headline:
Meet ARGUS, the robot built to catch hackers and physical intruders Read More »
When loading a model means loading an attacker 2025-10-03 at 10:15 By Sinisa Markovic You probably think twice before downloading a random app or opening an unfamiliar email attachment. But how often do you stop to consider what happens when your team downloads and loads a machine learning model? A recent study shows why you
React to this headline:
When loading a model means loading an attacker Read More »
The energy sector is ground zero for global cyber activity 2025-10-02 at 08:27 By Sinisa Markovic A new study from the Karlsruhe Institute of Technology shows how geopolitical tensions shape cyberattacks on power grids, fuel systems, and other critical infrastructure. How the research was done Researchers reviewed major cyber threat databases including MITRE ATT&CK Groups,
React to this headline:
The energy sector is ground zero for global cyber activity Read More »
The hidden risks inside open-source code 2025-09-30 at 09:12 By Mirko Zorz Open-source software is everywhere. It runs the browsers we use, the apps we rely on, and the infrastructure that keeps businesses connected. For many security leaders, it is simply part of the environment, not something they think about every day. That is where
React to this headline:
The hidden risks inside open-source code Read More »
Predicting DDoS attacks: How deep learning could give defenders an early warning 2025-09-25 at 10:12 By Sinisa Markovic Distributed denial-of-service (DDoS) attacks remain one of the most common and disruptive forms of cybercrime. Defenders have traditionally focused on detecting these attacks once they are underway. New research suggests that predicting DDoS attacks in advance may
React to this headline:
Predicting DDoS attacks: How deep learning could give defenders an early warning Read More »
LLMs can boost cybersecurity decisions, but not for everyone 2025-09-19 at 09:11 By Mirko Zorz LLMs are moving fast from experimentation to daily use in cybersecurity. Teams are starting to use them to sort through threat intelligence, guide incident response, and help analysts handle repetitive work. But adding AI into the decision-making process brings new
React to this headline:
LLMs can boost cybersecurity decisions, but not for everyone Read More »
Researchers believe Gamaredon and Turla threat groups are collaborating 2025-09-19 at 08:31 By Help Net Security ESET Research has discovered evidence of collaboration between the Gamaredon and Turla threat groups. Both groups are linked to Russia’s primary intelligence agency, the FSB, and were found working in tandem to target high-profile organizations in Ukraine. In these
React to this headline:
Researchers believe Gamaredon and Turla threat groups are collaborating Read More »