Artificial intelligence (AI) is advancing at a rapid rate, bringing new security threats and safety concerns to the fore. It will be critical that organisations can develop and deploy secure AI applications. To meet that need, Cisco has unveiled Cisco AI Defense, an end-to-end solution that protects both the development and use of AI applications.
Jeetu Patel, executive vice president and chief product officer, Cisco, discussed how quickly AI has become part of our lives and what that means for businesses. He said, “If you take a look at the market, there’s soon only going to be two types of companies. There’s going to be companies that are AI forward, and then there’s going to be companies that will be deemed irrelevant.
“Over the past couple of years, we’ve made a tremendous amount of progress in artificial intelligence. You can start to see this in every single thing that we are doing on a daily basis. There’s a huge amount of innovation that’s happening, and what seemed like very futuristic talk is getting to be just around the corner.”
Unprecedented risks
One of those “futuristic” technologies is “super intelligence”. Patel explained, “Super intelligence, which is where the machine exceeds the collective cognitive capability of humans, is something that is in sight at this point in time. So, there is a fair amount of velocity and progression that’s happening now. As this progression continues, the workforce composition is going to change quite dramatically.
“[Today] the workforce is 100 per cent humans, but we will see a huge augmentation of this workforce, with AI agents, AI apps, robots, and humanoids [coming in]. This is going to make a population of 8 billion feel like it has the capacity and throughput of 80 billion. But this massive augmentation of capacity [will mean] there’s a whole new class of risks that will start emerging at unprecedented scale.”
Patel pointed to the 2024 AI Readiness Index, a recent study by Cisco, which found that 29 per cent of those surveyed felt fully equipped to detect and prevent unauthorised tampering with AI. That shows many are looking for support in protecting their AI applications.
Those AI applications are also becoming multi-model and multi-cloud, creating new and complex security challenges. This means that vulnerabilities can occur at model or app level, with further complexity created by responsibility lying with different owners including developers, end users and vendors. The risks will be heightened further once AI models are trained on proprietary data.
One set of guardrails
That is why, Patel explained, Cisco has created AI Defense. This new solution looks to provide a common layer of safety and security that protects every user and every application when adopting and using AI. Patel said this new solution will address two urgent risks: the development of secure AI applications, and their deployment.
“The goal is to make sure that organisations can innovate fearlessly and not have to worry about safety and security as the primary concern, which could hold them back from accelerated adoption,” said Patel.
Cisco expects that, as AI becomes ubiquitous, organisations will use and develop a huge number of AI applications. AI Defense will provide one set of AI security and safety guardrails for every application, helping to protect AI systems from attacks and safeguard the behaviour of the AI models in use.
Joint success
Cisco’s partner community will be able to use this solution to solve problems for their customers, and Patel noted many partners are already setting up their AI practices and figuring out how their customers can use AI effectively.
Patel added, “We are a partner-first company. We would not be who we are as a company, if it weren’t for our partners. We think about our partners every step along the way, so AI Defense will be enabled to our partners. In the first few months [after the launch], we’re going to be in the learning process.
“We have to see how organisations are going to consume this, because this is all new for everyone right now. So, we’ll be learning for the first few months, and then we’ll scale it out to partners globally.”
The story of AI is still being written, but Patel is optimistic about the role this new product can play. He commented, “We’ve been deeply thinking about this for about a year, and I can’t tell you how excited I am that this product is coming to life. There’s going to be a meaningful impact to customers by using this product and accelerating their adoption [of AI].”
Cisco AI Defense will be available in March.
This interview was included in our February 2025 print issue. You can read the magazine in full here.