In the race to gain a competitive edge, organizations are increasingly training artificial intelligence (AI) models on sensitive data. But what if a seemingly harmless AI model became a gateway for attackers?
A malicious actor could upload a poisoned model to a public repository, and without realizing it, your team could deploy it in your environment. Once active, that model could exfiltrate your sensitive machine learning (ML) models and fine-tuned large language model (LLM) adapters. With access to these adapters, attackers could replicate your custom tuning and optimizations, exposing sensitive information embedded in fine-tuning patterns. Palo Alto Networks researchers recently uncovered two vulnerabilities in Google’s Vertex AI platform. These vulnerabilities could have allowed attackers to escalate privileges and exfiltrate models. We
Read more…
Source: Palo Alto Unit 42
Related:
- Have I Been Pwned claims Pitney Bowes hit by 8.2M email address leak
April 29, 2026
Logistics technology company Pitney Bowes, which makes franking machines for US postage, is the latest scalp claimed by ShinyHunters and its ongoing spree of pay-or-leak attacks against major organizations. Data breach tracker Have I Been Pwned (HIBP) confirmed the breach on April 27, with 8.2 million unique email addresses included in the dump alongside names, phone ...
- Medtronic says ShinyHunters hackers stole around 9 million medical records in latest attack
April 28, 2026
Medtronic, one of the biggest medical device manufacturers in the world, has confirmed suffering a cyberattack in which crooks “accessed data in certain Medtronic corporate IT systems.” In a security notification published on its website, Medtronic said the attack does not affect its customers or products, and also stressed it will continue operating as usual, without ...
- Don’t pay Vect a ransom – your data’s likely already wiped out
April 28, 2026
Organizations hit by the wave of Trivy and Lite LLM supply-chain compromises that paid Vect in hopes of recovering their data likely did not get much back, according to Check Point Research. That’s because the ransomware Vect uses isn’t actually ransomware at all, but a wiper that destroys any file larger than 128KB. Vect’s leak site ...
- Chinese engineer stole US military and NASA software for years
April 28, 2026
International espionage isn’t always about sophisticated malware and zero-day bugs. Sometimes it’s as simple as pretending to be someone else asking for a favor. For four years, a Chinese aerospace engineer did just that. Dozens of researchers at NASA, the US military, and major universities handed him exactly what he asked for, and possibly violated US ...
- ADT confirms cyber intrusion after ShinyHunters extortion attempt
April 27, 2026
A home security biz getting digitally burgled is not a great look – but that’s exactly where ADT finds itself. The company has confirmed a cyber intrusion following an extortion attempt by the ShinyHunters crew, which claims to have made off with more than 10 million records. US-based ADT is one of the world’s largest providers ...
- Attackers use hidden SMS and signalling systems to track targets’ location
April 24, 2026
Security researchers have just unveiled details of two covert surveillance campaigns that exploit weaknesses in the global telecom infrastructure. In a report published on Thursday, Citizen Lab explains that attackers abuse the signalling systems mobile operators use to support roaming, route messages, and locate devices on the network. The weaknesses were used to track certain subscribers ...