Source: CyberScoop
Researchers from NYU have unveiled ‘PromptLock’, an AI-powered malware that underscores the vulnerabilities in contemporary AI systems. Despite assurances from AI companies about robust safety measures, recent findings reveal that popular AI models are susceptible to prompt attacks. These models lack essential security features, making them easy targets for cybercriminals. As regulatory priorities shift, the need for effective AI security is becoming more critical than ever.
👉 Pročitaj original: CyberScoop
