New Zero-Click Attack Lets ChatGPT User Steal Data
Researchers at Radware discovered new prompt injection attacks in ChatGPT agentic features
Theft in cybersecurity covers stolen data, credentials, devices, and funds, often creating risks of unauthorized access, fraud, and privacy loss.
Search across headline titles and summaries.
Background for this topic.
Unauthorized taking or copying of information, credentials, intellectual property, or digital assets is cyber theft. News under this tag may involve stolen passwords, payment data, personal information, source code, cloud tokens, cryptocurrency, or sensitive business files. Theft can result from phishing, malware, compromised accounts, insider access, exposed storage, or the loss of an unencrypted device; the relevant issue is the unauthorized acquisition or control of an asset, whether or not the attacker also alters systems.
Security teams should identify where valuable data and credentials are stored, restrict access by role, require strong authentication, encrypt data at rest and in transit, and monitor unusual downloads or transfers. Vulnerability management matters when flaws expose databases, endpoints, or cloud services to unauthorized retrieval. After suspected theft, preserving logs, revoking tokens and credentials, determining what was accessed or copied, and assessing privacy or notification obligations are central to containing the incident and measuring its impact.
Researchers at Radware discovered new prompt injection attacks in ChatGPT agentic features
A new Python-based malware called VVS stealer has been identified, targeting Discord users with stealthy techniques to steal data
TRM Labs says it has recorded $35m drained from users’ wallets following 2022 LastPass breach