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:
- Horns&Hooves campaign delivers NetSupport RAT and BurnsRAT
December 2, 2024
Recent months have seen a surge in mailings with lookalike email attachments in the form of a ZIP archive containing JScript scripts. The script files – disguised as requests and bids from potential customers or partners – bear names such as “Запрос цены и предложения от Индивидуального предпринимателя <ФИО> на август 2024. According to Kaspersky telemetry, ...
- Retail outages drag into second week after Blue Yonder ransomware attack
December 2, 2024
A ransomware attack on supply chain software giant Blue Yonder continues to cause disruption to the company’s customers, almost two weeks after the outage first began. In a brief update to its cybersecurity incident page on Sunday, Arizona-based Blue Yonder said it is making “good progress” in its recovery from the attack, which hit its manage ...
- Shin Bet finds 200 Iranian cyberattacks on Israeli personalities
December 2, 2024
In recent months, the Shin Bet (Israel Security Agency) has uncovered some 200 efforts made by Iranian hackers to target Israeli civilians, the Shin Bet stated on Monday. The hacking was conducted via phishing attempts against various individuals, including Israeli politicians, academics, and media personalities, the security agency added. The hackers reportedly looked to gain access ...
- No company too small for Phobos ransomware gang, indictment reveals
December 2, 2024
The US Department of Justice has charged a Russian national named Evgenii Ptitsyn with selling, operating, and distributing a ransomware variant known as “Phobos” during a four-year cybercriminal campaign that extorted at least $16 million from victims across the world. The government’s indictment against Ptitsyn should dispel any notion that ransomware gangs only target the largest, ...
- RansomHub claims to net data hat-trick against Bologna FC
November 30, 2024
Italian professional football club Bologna FC is allegedly a recent victim of the RansomHub cybercrime gang, according to the group’s dark web postings. The ransomware crims responsible for attacks on organizations including Planned Parenthood and Christie’s – the same crew thought to have picked up LockBit’s top talent post-disruption – posted an extensive collection of data ...
- Some London commuters may never be refunded after TfL cyber attack
November 30, 2024
Sadiq Khan has admitted victims of a Transport for London (TfL) cyber attack may never get their money back. Tens of thousands of Londoners are feared to have been left out of pocket after hackers gained access to the travel authority’s systems in September. The aftermath of the hack meant over-60s, children and students were unable to ...