Hi everyone, in this video, we will explore Cisco's AI defense. AI defense is a comprehensive purpose built AI security solution that mitigates the risks associated with the development and deployment of AI models, applications, and agents.00:04AI defense empowers your security teams to discover AI assets across distributed cloud environments, detect vulnerabilities through supply chain scanning and algorithmic testing and protect runtime applications with security, privacy, and safety guardrails updated with the latest threat intelligence. Let's get started.00:19With the rise of AI, new security threats emerge. AI applications are fundamentally different than traditional software applications. Generative models and agents introduce non deterministic elements into the application stack, meaning the same input can result in different outputs.00:39This creates new safety and security risks vectors that aren't addressed by existing cybersecurity solutions such as unsafe or inappropriate responses, data exfiltration or leakage, failure to comply with AI security standards and regulations.00:56AI defenses framework for AI security has three steps. Discover what AI assets exist in your business, detect their risks and vulnerabilities, and protect them with guardrails that cover these gaps. AI defense provides protection across the AI lifecycle.01:12AI inventory provides an inventory of all AI assets, inventory, the AI models, agents, and connected data sources across distributed environment to understand usage and gauge risk. AI supply chain risk management scans open source components for security threats.01:29Scan model files, repos and MCP servers to proactively block malicious or unsafe AI assets before operations are impacted. AI model and application validation uses algorithmic red teaming to test for vulnerabilities, identify safety and security vulnerabilities across models at scale with algorithmic red teaming technology.01:45AI runtime protection secures applications in production, protect production all apps and agents with guardrails embedded in the network, block attacks and harmful responses in real time. AI inventory provides a foundational understanding of what AI assets exist within your cloud environment and externally, automatically uncovered.02:05AI assets across your distributed cloud environment, including models, agents, and connected data sources, understand important usage context around AI assets. AI supply chain risk management scans 3rd party assets like open source models to identify vulnerabilities.02:25This ensures you can build AI applications with components that aren't compromised, automatically scan model files in your private repositories to identify vulnerabilities like code execution and suspicious imports scan MCP servers to inventory tools and detect tool poisoning attacks and prevent the usage of insecure models and 3rd party assets.02:43Next, you want to identify vulnerabilities in your AI models and applications. With algorithmic red teaming technology, our product accomplishes in minutes what would otherwise take weeks of manual testing. Identify vulnerabilities in models and applications through automated algorithmic AI red teaming automatically generate reports that matter.03:04To AI security standards such as a WASP and MITRE atlas.03:22AI defense allows you to perform risk focused validation with multi turn validation and custom goals. Algorithmic red teaming automatically and continuously scans across 200 plus safety and security subcategories, testing your model to see where the risks are. You then get a report with recommendations on guardrails to mitigate those risks.03:27AI defense's risk focused validation process allows advanced multi turn validation which simulate realistic conversational attack scenarios that evolve over multiple turns, allowing you to test your AI application's resilience against sophisticated adaptive threats that go beyond single prompt interactions.03:47As well as to execute custom attack goals through both the UI and API enabling you to tailor validation tests to your unique business risks and target endpoints. By leveraging validation, customer enhance their security posture with a risk centric interface that delivers actionable risk focused insights, granular technique breakdowns.04:04Automated risk segmentation and quantify performance metrics. Now you know what AI assets exist in your environment and what they're vulnerable to. AI runtime protection is the final piece, providing guardrails that cover your models against the risks you previously identified.04:23Define bi directional guardrails for applications and agents that block malicious prompts and unsafe responses. Customize them to cover specific model vulnerabilities and fit unique AI applications. The guardrails provided by AI defessents runtime protection have broad coverage and are subject to ongoing updates to protect against emerging threats.04:40They're divided into three main categories, security guardrails block attempts to override an llm's alignment, generate or retrieve code, maximize resource consumption or execute other insecure outcomes like prompt injection, code presence, cybersecurity and hacking, adversarial content or tool misuse.05:00Privacy guardrails block attempts to reveal sensitive information that might cause harm if lead to an unauthorized party, such as intellectual property, IP, theft, sensitive data disclosure, including PII, PHI, PCI, metaprompt extraction or exfiltration from AI application.05:18Safety guardrails prevent the generation of content which may offend or cause harm to individuals, businesses and broader society. Hate speech and profanity, sexual content, harassment, violence, and public safety threats, and rogue agents. By integrating AI defense within their workflow.05:36AI dev teams will seamlessly extend security into all stages of the development life cycle. This promotes a shift left approach to AI security, incorporating important measures like supply chain scanning and model validation earlier on in AI development. AI defense integration with other business critical tools help customers operation.05:55Analyze AI security.06:16Splunk enterprise security to correlate AI telemetry and create alerts, ServiceNow AI control tower to operationalize AI security. Safe security will help quantify and manage AI risk. AI Defense offers the convenience of a Cisco hosted cloud managed control plane with the flexibility of deployment options.06:17For the data plane.06:37SAS VPC or AI pods. SaaS for customers looking for a simple, flexible deployment with zero infrastructure to manage. VPC for customers looking to balance data control and compliance with cloud scalability, and the customers that wanna manage AI workloads themselves rather than relying on hyperscalers, will be able to do it on their own data.06:38Data center. By integrating Cisco AI defenses, discover, detect, and protect use cases into your AI strategy, you equip your organization to push the boundaries of innovation safely, swiftly, and with complete confidence in your security risk posture. Thank you for watching and make sure you check the other AI defense cyber capsules.06:58
