The rapid evolution of artificial intelligence (AI) presents exciting possibilities but also demands a strong focus on security. At Intel, the decided to take these challenges head on teaming with developers and adopters in safeguarding AI technology while navigating a complex landscape of guidelines and standards. Intel, along with other industry giants, believe in the power of collective action to address these challenges and ensure that security remains a top priority.
Intel Joins Forces with the Coalition for Secure AI (CoSAI)
Intel's CEO, Pat Gelsinger, recently unveiled a strategy for open and scalable AI systems at Intel Vision 2024. To further this vision, Intel joined the Coalition for Secure AI (CoSAI) as a founding member. CoSAI aims to equip practitioners and developers with the tools and guidance needed to build secure AI systems from the ground up.
This collaboration is a significant step for the industry, uniting leaders from diverse fields to develop and share comprehensive approaches, best practices, and methodologies for secure AI development and deployment. CoSAI's initial focus will be on software supply chain security, preparing defenders for evolving cybersecurity threats, and establishing AI security governance frameworks.
Prioritizing Security and User Protection in the AI Boom
The demand for AI is skyrocketing, and so are the security risks. As we have seen this week with the CrowdStrike outage cybersecurity is still evolving to keep up. It would seem with the widespread impact of thi outage that demand is far outpacing the current capabilities of some vendors and cybersecurity systems and experts. CoSAI hopes that with a focus on protecting AI systems from attacks, ensuring their transparency, and building trust with users they will be able to prevent future situations like we have today.
Unfortunately, we are already seeing hackers utilizing AI systems in taking advantage of the CrowdStrike outage chaos. They are setting up AI generated websites and portals, using realistic AI generated email for spear-phishing attacks and as CrowdStrike itself warned on their own blog post hackers are even impersonating CrowdStrike employees in scam emails and phone calls, even selling bogus software purporting to fix the glitch.
As AI systems evolve and become even more prevalent, we are going to need groups like the CoSAI working with teams from Intel, NVIDIA, Microsoft, Google and others to do their due diligence!