Geek-Guy.com

Category: Research

Auto Added by WPeMatico

These researchers used NPR Sunday Puzzle questions to benchmark AI ‘reasoning’ models

Every Sunday, NPR host Will Shortz, The New York Times’ crossword puzzle guru, gets to quiz thousands of listeners in a long-running segment called the Sunday Puzzle. While written to be solvable without too much foreknowledge, the brainteasers are usually challenging even for skilled contestants. That’s why some experts think they’re a promising way to…

Threat researchers spot ‘device code’ phishing attacks targeting Microsoft accounts

Microsoft threat researchers discovered a series of what they are calling “device code” phishing attacks that allowed a suspected Russia-aligned threat group to gain access to and steal data from critical infrastructure organizations, the company said in research released Thursday. The group, which Microsoft tracks as Storm-2372, has targeted governments, IT services and organizations operating…

Salt Typhoon remains active, hits more telecom networks via Cisco routers

Salt Typhoon, the Chinese nation-state threat group linked to a spree of attacks on U.S. and global telecom providers, remains active in its intrusion and has hit multiple additional networks worldwide, including two in the United States, Recorded Future said in a report released Thursday. Recorded Future’s Insikt Group observed seven compromised Cisco network devices communicating…

Russian state threat group shifts focus to US, UK targets

A subgroup of Seashell Blizzard has shifted its focus to targets in the U.S., Canada, Australia and the U.K. within the past year, expanding the scope of its malicious activity, Microsoft’s threat intelligence team said in a report released Wednesday. The initial-access operation, which Microsoft tracks as the “BadPilot campaign,” has allowed the Russian state…

ChatGPT may not be as power-hungry as once assumed

ChatGPT, OpenAI’s chatbot platform, may not be as power-hungry as once assumed. But its appetite largely depends on how ChatGPT is being used, and the AI models that are answering the queries, according to a new study. A recent analysis by Epoch AI, a nonprofit AI research institute, attempted to calculate how much energy a…

Is AI making us dumb?

Researchers from Microsoft and Carnegie Mellon University recently published a study looking at how using generative AI at work affects critical thinking skills. “Used improperly, technologies can and do result in the deterioration of cognitive faculties that ought to be preserved,” the paper states. When people rely on generative AI at work, their effort shifts…

These researchers used NPR Sunday Puzzle questions to benchmark AI ‘reasoning’ models

Every Sunday, NPR host Will Shortz, The New York Times’ crossword puzzle guru, gets to quiz thousands of listeners in a long-running segment called the Sunday Puzzle. While written to be solvable without too much foreknowledge, the brainteasers are usually challenging even for skilled contestants. That’s why some experts think they’re a promising way to…

Infosec pros: We need CVSS, warts and all

A key pillar of a strong cybersecurity program is identifying vulnerabilities in the complex mix of software programs, packages, apps, and snippets driving all activities across an organization’s digital infrastructure. At the heart of spotting and fixing these flaws is the widely used Common Vulnerability Scoring System (CVSS), maintained by a nonprofit called the Forum…

Here’s all the ways an abandoned cloud instance can cause security issues

There is a line of thought among the public that “the internet is forever.” A security company published research Tuesday that showed why “forever” can be a security nightmare.  Over the course of four months, cybersecurity researchers at watchTowr monitored and ultimately took control of what they referred to as “abandoned” digital infrastructure, focusing on…

From credit card fraud to zero-day exploits: Xe Group expanding cybercriminal efforts

A cybercriminal organization that has been operating for over a decade has moved from credit-card skimming to exploiting zero-day vulnerabilities, according to a joint investigation by cybersecurity firms Solis Security and Intezer. The group, tracked as XE Group, now poses heightened risks to global supply chains, particularly in manufacturing and distribution sectors, by leveraging stealthier…

Hackers Hijack JFK File Release: Malware & Phishing Surge

Veriti Research has uncovered a potentially growing cyber threat campaign surrounding the release of the declassified JFK, RFK, and MLK files. Attackers are capitalizing on public interest in these historical documents to launch potential malware campaigns, phishing schemes, and exploit attempts. Our research indicates that cybercriminals are quick to react to major public events, and…

DARPA wants to create ‘self-healing’ firmware that can respond and recover from cyberattacks 

Imagine, for a moment, that your network is hit with ransomware. One of your employees clicked on a malicious link and now your network is compromised, data is encrypted and most of the organization’s systems are locked or offline. Then imagine if instead of assembling an incident response team,  notifying the board and contacting law…

Vulnerability in popular AI developer could ‘shut down essentially everything you own’ 

A popular platform for developing AI systems has patched an easily exploitable vulnerability that would have given an attacker remote code execution privileges. Researchers at application security firm Noma detail how the flaw, embedded in Javascript code for Lightning.AI’s development platform, could be manipulated to give an attacker virtually unfettered access to a user’s cloud…

Open-source security spat leads companies to join forces for new tool

A conflux of open-source developers and application security companies has been embroiled in a complex debate after a recent change in the licensing policy of a widely used static code analysis tool, resulting in a faction of organizations creating a new, open-source rival.  The issue started with a recent change in the licensing policy of…

The AI industry’s pace has researchers stressed

To outside observers, AI researchers are in an enviable position. They’re sought after by tech giants. They’re taking home eye-popping salaries. And they’re in the hottest industry of the moment. But all this comes with intense pressure. More than half a dozen researchers TechCrunch spoke with, some of whom requested anonymity for fear of reprisals,…

Japanese Companies Threatened by DPRK IT Workers

Nisos Japanese Companies Threatened by DPRK IT Workers The Japanese government warned domestic companies in March 2024 about contracting North Korean (DPRK) IT workers posing as Japanese nationals to earn cash, as it is suspected… The post Japanese Companies Threatened by DPRK IT Workers appeared first on Nisos by Nisos The post Japanese Companies Threatened…

‘Severe’ bug in ChatGPT’s API could be used to DDoS websites

A vulnerability in ChatGPT’s API can generate DDoS attacks against targeted websites, but the security researcher who discovered it says the flaw has since been addressed by OpenAI. In a security advisory posted to the developer platform GitHub, German security researcher Benjamin Flesch detailed the bug, which occurs when the API is processing HTTP POST…

CloudFlare detected (and blocked) the biggest DDoS attack on record

Web infrastructure and security company Cloudflare says it detected the biggest Distributed Denial-of-Service (DDoS) attack ever recorded, a 5.6 terabits per second (Tbps) attack directed at an internet service provider (ISP) in Eastern Asia. Despite the staggering volume of the attack, Cloudflare successfully managed and mitigated it without human intervention. The company said in research…

Trump’s Digital Footprint: Unveiling Malicious Campaigns Amid Political Milestones

As the political landscape heats up, so does the activity in the cyber threat domain. High-profile events such as inaugurations often become a prime opportunity for cybercriminals to launch malicious campaigns. With Trump’s upcoming inauguration on January 20th, our research sheds light on the digital threats tied to such politically charged events, focusing on previous…

The Insider Threat Digital Recruitment Marketplace

Nisos The Insider Threat Digital Recruitment Marketplace Nisos routinely monitors mainstream and alternative social media platforms, as well as cloud-based messaging applications and dark web forums… The post The Insider Threat Digital Recruitment Marketplace appeared first on Nisos by Nisos The post The Insider Threat Digital Recruitment Marketplace appeared first on Security Boulevard.

Researchers open source Sky-T1, a ‘reasoning’ AI model that can be trained for less than $450

So-called reasoning AI models are becoming easier — and cheaper — to develop. On Friday, NovaSky, a team of researchers based out of UC Berkeley’s Sky Computing Lab, released Sky-T1-32B-Preview, a reasoning model that’s competitive with an earlier version of OpenAI’s o1 on a number of key benchmarks. Sky-T1 appears to be the first truly…

A popular technique to make AI more efficient has drawbacks

One of the most widely used techniques to make AI models more efficient, quantization, has limits — and the industry could be fast approaching them. In the context of AI, quantization refers to lowering the number of bits — the smallest units a computer can process — needed to represent information. Consider this analogy: When…

New Anthropic study shows AI really doesn’t want to be forced to change its views

AI models can deceive, new research from Anthropic shows — pretending to have different views during training when in reality maintaining their original preferences. There’s no reason for panic now, the team behind the study said. Yet they said their work could be critical in understanding potential threats from future, more capable AI systems. “Our…

‘Reasoning’ AI models have become a trend, for better or worse

Call it a reasoning renaissance. In the wake of the release of OpenAI’s o1, a so-called reasoning model, there’s been an explosion of reasoning models from rival AI labs. In early November, DeepSeek, an AI research company funded by quantitative traders, launched a preview of its first reasoning algorithm, DeepSeek-R1. That same month, Alibaba’s Qwen…

Microsoft debuts Phi-4, a new generative AI model, in research preview

Microsoft has announced the newest addition to its Phi family of generative AI models. Called Phi-4, the model is improved in several areas over its predecessors, Microsoft claims — in particular math problem solving. That’s partly the result of improved training data quality. Phi-4 is available in very limited access as of Thursday night: only…

Turla living off other cybercriminals’ tools in order to attack Ukrainian targets

A Russian nation-state threat actor has been observed leveraging tools from other cybercriminal groups to compromise targets in Ukraine, a recent report by Microsoft Threat Intelligence disclosed. This clandestine approach, which is the second time in as many weeks that Microsoft has highlighted the group’s effort, shows how Turla uses a wide range of attack…

Latest round of MITRE ATT&CK evaluations put cybersecurity products through rigors of ransomware 

MITRE Corporation released findings Wednesday from its latest round of ATT&CK evaluations, assessing the capabilities of enterprise cybersecurity solutions against some of the most prevalent ransomware tactics and North Korean malware. The sixth such evaluation from the nonprofit research organization measured 19 different vendors’ ability to protect enterprise systems by evaluating them against two prominent…

WordPress Appliance - Powered by TurnKey Linux