Analysis of Google finds ‘LLMjacking’ attackers running up $100K a day in fraudulent cloud charges - aVenture Research
SiliconANGLEby Duncan Riley
Google finds ‘LLMjacking’ attackers running up $100K a day in fraudulent cloud charges
A new report out today from Google LLC’s Mandiant and Google Threat Intelligence Group finds that attackers who hijack enterprise access to large language models are running up fraudulent cloud charges of more than $100,000 in a single day. The report, “The Rise of LLMJacking: How Threat Actors Mone
This is an account of investigated cases, not a measure of prevalence across cloud customers. The immediate business implication is that AI-service access controls and billing safeguards are also controls against direct cloud-cost shock; the article’s decisive uncertainty is how frequently attackers can sustain access before organizations detect and revoke credentials.
Google’s security teams are the source of the investigation and threat assessment, rather than a party shown to have suffered the reported charges. For Google, that establishes a security and trust exposure around its AI service, but the article supplies no evidence of a direct financial loss to Google or a quantified effect on Gemini adoption. The practical stake supported here is whether customer safeguards and incident response can contain misuse without undermining legitimate enterprise use; the article reports recommendations, not their measured effectiveness.