Foundation AI: Open-Source Innovations for Secure AI
To address new and complex cybersecurity challenges introduced by the rapid adoption of artificial intelligence (AI), Cisco launches Foundation AI, a team of leading experts in AI and cybersecurity. This group is fully dedicated to solving major security challenges of the AI era by developing advanced tools and technologies that tackle these core issues.
This new approach reflects the pressing need to balance rapid AI adoption with robust security measures. Foundation AI’s tools will not only empower organizations to defend against emerging threats but will also support policy goals to create ethical and secure AI systems.
The Foundation AI team has just released the first-ever open-source reasoning model designed specifically for security purposes. They are also introducing benchmarks to test how well cybersecurity models perform in real-world scenarios, as well as tools that teams can leverage to customize and improve their own models. These efforts will foster collaboration between security experts, machine learning engineers, and AI developers, providing practical solutions that businesses can immediately leverage to improve their cybersecurity systems.
The open-source tools and benchmarks launched by Foundation AI support the EU’s goals to foster collaboration and transparency in technology development. By encouraging the use of secure and ethical AI, Cisco’s initiatives contribute to the EU’s vision of a digital ecosystem that prioritizes safety, resilience, and innovation, for businesses and society at large.
AI Supply Chain Risk Management: Stopping Malicious or Risky AI Before Harm
Cisco is enhancing its AI Supply Chain Risk Management capabilities to help organizations tackle the growing challenges of managing AI security risks and safely adopt and innovate with AI. These tools protect organizations by identifying and stopping malicious or risky AI models before they can cause harm. For instance, they can:
- automatically detect and block AI models with risky or restrictive open-source software licenses that pose intellectual property and compliance risks;
- enforce policies on AI models originating from geopolitically sensitive regions;
- detect and prevent the use of harmful AI models in the organization’s environment.
These advancements are part of a larger vision to integrate security at every stage of AI adoption, safeguarding organizations in the rapidly evolving AI landscape.
Secure AI Deployment and Resilience
With the AI Act, the EU provided a framework to address risks associated with AI and promote its trustworthy use, with an emphasis on making high-risk AI systems accurate, robust and secure. Cisco’s approach and technologies can help deployers and users of these systems make sure that is the case.
This focus also comes as the EU positions itself in the global AI landscape, with significant investments aimed at creating an AI-ready continent. Tools like those introduced by Cisco’s Foundation AI enhance the resilience of AI deployment across Europe, aligning with the EU’s ambition to protect critical infrastructure, ensure compliance, and reinforce cybersecurity.
As AI adoption accelerates, these efforts exemplify the need for a balanced approach—one that embraces innovation while safeguarding against emerging risks. By building practical solutions, initiatives like Foundation AI not only strengthen security but also advance the collective vision of a secure and ethical AI future.
Share:
