Tech giant Cisco said it has unveiled a new security solution designed to protect artificial intelligence systems within enterprise environments, as companies grapple with emerging safety concerns around AI deployment.
The new platform, called Cisco AI Defense, aims to address two key areas of risk: securing AI application development and protecting access to AI applications.
Recent research from Cisco's 2024 AI Readiness Index revealed that only 29 per cent of organisations felt fully prepared to detect and prevent unauthorised tampering with AI systems.
"Business and technology leaders can't afford to sacrifice safety for speed when embracing AI," Jeetu Patel, chief product officer, Cisco, said.
The solution includes features for discovering AI applications across public and private clouds, validating AI models for potential safety issues, and implementing runtime security to protect against threats like prompt injection and data leakage.
For end users, the platform provides visibility of AI-enabled applications, controls employee access to unsanctioned AI tools, and protects against confidential data loss.
Cisco has developed the system to work with multiple AI models simultaneously, leveraging its proprietary machine learning capabilities and threat intelligence data from Cisco Talos, its cybersecurity research specialists.
The company has worked alongside industry standard-setting organisations including MITRE, OWASP, and NIST to ensure the platform meets recognised security benchmarks.
The new security solution will be available to enterprise customers from March 2024.
-1.jpg&w=100&c=1&s=0)
.png&w=100&c=1&s=0)
