Plainly speaking, Artificial intelligence is no longer a fringe technology. It has become a core component of modern business, from customer service chatbots to complex data analysis.
We often treat the Large Language Models (LLMs) that are at the core of this technology as trusted black boxes. But like any software, they can be tampered with, manipulated, and turned against their creators. Understanding the ways an AI model can be compromised is the first step toward building a secure and resilient AI infrastructure. There are three primary ways that the authors want to cover, in which they will explain how an AI model can be compromised and made to act in unpredictable (or malicious!) ways:
Read more…
Source: Trend Micro
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- How to stop fake System notifications on macOS
November 21, 2023
Scammers are abusing an Apple feature that allows websites to create push notifications that look like they’re coming from macOS, or apps. The notifications try to scare users into clicking a link with fake virus alerts or messages saying their account has been hacked. Years ago Malwarebytes Labs warned our readers about the introduction of browser ...
- British Library Employee data leaked in cyber attack
November 21, 2023
The British Library has confirmed that a cyber attack in October has led to a leak of employee data. The attack, which took place on 31 October, has also resulted in the library’s website being down for almost a month. The Rhysida ransomware group claim to be behind the attack, and say they will auction off ...
- Hacking Employers and Seeking Employment: Two Job-Related Campaigns Bear Hallmarks of North Korean Threat Actors
November 21, 2023
Unit 42 researchers recently discovered two separate campaigns targeting job-seeking activities linked to state-sponsored threat actors associated with the Democratic People’s Republic of Korea (DPRK), commonly known as North Korea. The research team call the first campaign “Contagious Interview,” where threat actors pose as employers (often anonymously or with vague identities) to lure software developers into ...
- U.S. DOD strategy warns emerging tech is ‘at the forefront’ of information threats
November 21, 2023
The Pentagon publicly released its strategy for operating in the information environment – which covers both physical and digital sources of information – on Friday, outlining how the agency plans to modernize its collecting, processing and sharing of data to better counteract adversaries’ weaponization of the internet and emerging technologies. DOD “must embrace a cultural shift ...
- SysAid path traversal vulnerability
November 21, 2023
SonicWall Capture Labs Threat Research Team became aware of the SysAid path traversal vulnerability, assessed its impact and developed mitigation measures for the vulnerability. On November 8, 2023, SysAid, an IT service management company, disclosed CVE-2023-47426, which is a zero-day path traversal vulnerability carrying a CVSS 9.8 score and affecting on-premise SysAid servers running version < ...
- #StopRansomware: LockBit 3.0 Ransomware Affiliates Exploit CVE 2023-4966 Citrix Bleed Vulnerability
November 21, 2023
The Cybersecurity and Infrastructure Security Agency (CISA), Federal Bureau of Investigation (FBI), Multi-State Information Sharing & Analysis Center (MS-ISAC), and Australian Signals Directorate’s Australian Cyber Security Center (ASD’s ACSC) are releasing this joint Cybersecurity Advisory (CSA) to disseminate IOCs, TTPs, and detection methods associated with LockBit 3.0 ransomware exploiting CVE-2023-4966, labeled Citrix Bleed, affecting Citrix ...