Cisco has opened up to the latest platform, January 15, 2025. This solution is developing security risk associated with counterfeit environments. The platforms focused on both development and use of the AI application to reduce weakness with AI systems.
[Read More: UTAR PhD Student Wins Cisco AI Hackathon with Anti-Procrastination AI Platform]
Important features of Cisco Ai Prevention
Cisco Ai defense mentioned two main concerns:
-
Ensure AI application development: AIF: Ai checkout to determine the weakness, such as the risk of toxic input or input, strong application.
-
Protection of access to AI application: It enhances monitoring by discovering and monitoring using AI while using the risk of unauthorized or leak tools.
Important features include:
-
AI Modelication: Auto processes to assess the weaknesses that may make.
-
Running time protection: Real-time monitoring to prevent threats like a shot and make profile data.
-
Discovering the application: AI warehouse tool and sources for improving visibility.
-
Access Control: The measure that will limit the use of third-party unauthorized AI programs and prevent sensitive information.
[Read More: Google Enhances Android Security with AI-Driven Scam Detection and Real-Time App Protection]
Collaboration and expertise on the platform
Cisco Ai learning technology learning technology and understanding from Cisco Talos, Ciscco Research Apps. By drawing in the vast research, the platform offers solutions that are aiming to address both current and emerging threats. Cisco also worked with grinding organizations such as miter, owak, and nist to organize the best performance.
The vast context of safety AI
As Ai becomes a business operation, it introduces new weaknesses at the traditional security system is not installed. Examples of these risks include:
-
Inverse attack models: Unabled attacks that can prevent harmful actors can extract sensitive information from the engine used during the engine. It means that this attacker can disclose confidential details such as customer data, or a myth of the owner, or a marital. The violation creates serious risks in the industry as well as healthy, where the information of personal patients.
-
Toxicity of toxicity: The poisonous attacks of the attackers that make the attacker denied harmful or accusatory information in the form of model. This manipulation can result in incorrect or dangerous results, destroying the credibility of AI system. For example, in the independent vehicles, poisoned a series may lead to false material detection, detained. Likewise, in financial decisions, such attacks can mention the estimates, leading to the system with the system.
-
Opposite example: The opposite example will be a variety of AI models by gentle recommendations, almost undispted, misleading form. Classic examples are scamming the vision system of cars that drive themselves to reading a speed sign, which can lead to an accident. These attacks indicate the active features of the AI threat and insufficarity of traditional defense, elimination of the risk.
These challenges highlight the need for dedicated solution in the uniform AI systems.
Compare CISCO with industrial leaders
Other companies also mentions AI security challenge, offering a solution to form or compete with Cisco methods:
-
IBM: IBM QRADAR SIEM uses AI to monitor network traffic, rational events and reminders to help the threatening agency. IBM keeper, with the ability to produce risky AI and management of the AI moderation and detective use of AI. These tools provide strong control services and compliance coverage.
-
NVIDIA: The Morgeus scope of the NVIDIA’s large process of real-time data using the GPU, making the developer for a threat. Tools such as Bluefield DPus and Doca Sdk helps to safeguard the safety zero, provide ability to check and authentication.
While IBM and NVIDIA offers strong solutions, Cisco Ai defense discrimination, AI protection, and uncovering for enterprises.
Availability and consequences of the industry
Cisco Ai defense will be available to enterprise to start on March 2025. As the organization is demanded on the target. By concentration of specific risk AI, CISCO platforms contributed to a safe and responsible effort.
The development of the tool emphasizes the importance to grow in solving the risk of fasting AI in fast technology. Organizations must be careful and invest in solutions that guarantee their AI systems while innocent.
Source: Cisco Press Room, Dryer, IBM, RS Studio, Approximate, Sentinel One, NVIDIA