In an exclusive interview on Sunday with Rowan Chung from The Rundown AI, Cisco's Executive Vice President and CPO, Jitu Patel, stated that AI Defense "takes a radical approach to solving problems that existing security solutions fail to address."
AI Defense, announced last week, aims to minimize risks during the development and deployment of artificial intelligence applications and to identify where AI is being used within an organization.
AI Defense Capabilities
The AI Defense system can protect AI systems from attacks and ensure secure model operations across various platforms through the following features:
- Detection of both shadow and authorized AI applications in public and private clouds.
- Automated AI model testing against hundreds of potential security vulnerabilities.
- Continuous monitoring to safeguard against threats such as rapid injection attacks, denial-of-service, and confidential data leaks.
The solution also enables security teams to enhance corporate data protection by providing full visibility into employees’ AI application usage, creating access policies for unauthorized tools, and implementing security measures to prevent confidential information leaks.
A Significant Step for AI Security
"The implementation of AI introduces new risks that traditional cybersecurity solutions fail to address," noted Kent Noyes, Global Head of AI and Cyber Innovation at World Wide Technology (St. Louis). "Cisco AI Defense is a major leap forward in AI security, providing complete visibility into enterprise assets and protection against the latest threats."
M. Jay Kaufmann, an author and instructor at O’Reilly Media, agrees with Cisco’s analysis of weaknesses in current cybersecurity solutions. "Cisco is right," she said in an interview with TechNewsWorld. "Existing tools fail to address many threats to AI systems, such as rapid injection attacks, data leaks, and unauthorized model actions."
"Organizations must take action and implement specialized solutions," she added.
Cisco Has Unique Capabilities to Deploy AI Defense
Jack E. Gold, founder and chief analyst at J.Gold Associates, pointed out that Cisco is uniquely positioned in this field. "They have vast amounts of data from their network telemetry, which can be leveraged to enhance AI security capabilities," he said in an interview with TechNewsWorld.
He also emphasized that Cisco is committed to securing AI across multiple platforms—on-premises, cloud, and multi-cloud—as well as different AI models.
"It will be interesting to see how many companies adopt this solution," Gold added. "Cisco is moving in the right direction, as most companies are still struggling to effectively address these issues."
Multi-Model, Multi-Cloud Protection: A Key Component of AI Security
"Multi-model, multi-cloud AI solutions expand an organization’s attack surface, making security processes more complex due to fragmented environments, differences in security protocols, multiple data transmission points, and the difficulty of coordinating monitoring and incident response," explained Patricia Thain, CEO and co-founder of Private AI (Toronto), to TechNewsWorld.
Cisco AI Defense could become a key player in the new era of AI security, helping companies gain better control and protection over their AI systems.