AI is undoubtably a game-changer for cyber security professionals around the world. But for many organizations there remains a degree of uncertainty around how best to implement and utilize. Cisco is committed to helping its customers understand and get the most out of AI, which is why it has announce the launch of the Cisco AI Security Portal at aisecurity.cisco.com. Designed to provide users with a toolkit to navigate AI in security, the portal provides the latest best practices, resources, insights, and guidance on securing systems and applications. It includes several white papers authored by experts from many teams across Cisco and Robust Intelligence (now part of Cisco).

The Cisco portal contains an array of resources that will help professionals working within AI development, deployment, and security. The following figure includes a high-level glimpse of the best practices that you can find in the content highlighted in the portal:

AI Training Environment Security and AI/ML Ops

Users will find best practices for securing AI training environments and managing AI/ML operations. Meanwhile, they can also learn how to protect data and models during the training phase and implement robust operational processes to maintain security throughout the AI lifecycle.

Securing Vector Databases

The portal will enable professionals to better understand the unique security best practices to secure vector databases, which are crucial for handling high-dimensional data in AI applications. Get guidance on implementing security measures to protect sensitive information and ensure data integrity.

Securing AI Systems Lifecycle: AI/MLOps, CRISPML( Q), and OpenEoX

Allow your team easy exploration of strategies for securing AI systems throughout their lifecycle, covering methodologies like AI/MLOps, CRISP-ML(Q), and OpenEoX, offering insights into integrating security at every stage of AI development and deployment.

AI Security Reference Architectures

Users will also be able to access detailed reference architectures from Robust Intelligence that provide blueprints for building secure AI systems. These architectures serve as practical guides for implementing security controls and can be adapted to suit various organizational needs and technological environments.

Defining Artificial Intelligence Security Vulnerabilities

Finally, users can gain a deeper understanding of the definition of AI Security Vulnerabilities. Access the FAQ about the disclosure of security vulnerabilities found in AI systems and how to report any AI vulnerabilities to the Cisco Product Security Incident Response Team (PSIRT).

Cisco Research is committed to investigating innovative and emerging domains that are strategically significant to Cisco, with the aim of making an impact on business, technology, and society. The portal links to many of these research academic papers and provides continuous access to a wide array of research areas, such as artificial intelligence and machine learning, cybersecurity, quantum computing, networking, and distributed systems.

Cisco actively collaborates with academic researchers from leading universities around the globe, providing funding and resources to advance cutting-edge research and drive meaningful outcomes.

We invite you to explore the AI Security Portal and take advantage of the resources available. Whether you’re a seasoned cybersecurity, AI professional, or new to the field, the portal offers valuable insights to help you secure your AI implementations.

www.cisco.com/c/en_za