Cisco earmarks $1B for AI

　　Cisco is placing a big bet on artificial intelligence (AI) as its investment arm launches a $1 billion investment fund focused on expanding and developing AI technologies and working with tech partners.

　　Tied in, Cisco said it is making a range of strategic investments in various AI-facing startups, including Cohere, Mistral AI and Scale AI, among others.

　　Cisco, which introduced its AI investment fund in concert with this week's Cisco Live event in Las Vegas, said it has already committed nearly $200 million toward its targeted $1 billion total. Cisco is creating the fund, its first focused on AI, on the heels of its $28 billion acquisition of AI and cybersecurity specialist Splunk.

　　AI is still a nascent technology and market, but it "will be fundamentally changing every industry around the globe," Mark Patterson, Cisco EVP and chief strategy officer, declared Monday at a Cisco Live event with press and analysts.

　　Vijoy Pandey (far right), SVP, Outshift, at Cisco, is joined by Mark Patterson (far left), EVP and CSO, and Derek Idemoto, SVP of corporate development and Cisco Investments at Cisco Live 2024. (Source: Jeff Baumgartner/Light Reading)

　　The broad idea is to provide infrastructure to power AI technologies, including a focus on the enterprise market and areas such as cybersecurity, according to Patterson. Cisco is pursuing that "across the stack" by developing its own products, investing in AI companies and striking AI-focused partnerships, he noted.

　　"This is what the AI fund is for – to go bigger and go faster," said Derek Idemoto, SVP of corporate development and Cisco Investments. "Partnerships are going to be very, very key there."

　　Idemoto said later that AI is poised to "melt into the network … There's going to be a lot of net-new [business] created."

　　Cybersecurity calls for collaboration

　　Discussions at the event also focused on the cybersecurity aspects of AI.

　　The emergence of AI is driving up the complexity of cybersecurity, Jeetu Patel, Cisco's EVP and GM, security and collaboration, said in a session Monday afternoon. While network operators can use AI tools to protect their systems, attackers also can implement them.

　　"AI is changing everything," Patel said. At the same time, AI "can tip the scale in favor of the defenders," he said during Tuesday morning's Cisco Live keynote.

　　"Security is a data problem," Patel said. The core philosophy is to assume that an attacker has already infiltrated the system, so it's important to contain the attacker's movements as rapidly as possible and close up vulnerabilities that can be patched, he added.

　　But Cisco was pressed about competition in the security arena, particularly from hyperscalers. Patel said it's important for all parties to work together and exchange data and best practices towards a common security goal to "keep the bad folks at bay."

　　"You can't think of competition as a zero-sum game," Patel said, later mentioning Microsoft as an example of a company that collaborates with Cisco on cybersecurity issues.

　　"The true enemy is the adversary, not the competitor," he said.

　　Among other AI-themed announcements this week, Cisco, in tandem with Nvidia, introduced Nexus HyperFabric AI clusters, billing it as a simplified platform for data centers for generative AI applications.

　　The on-premises platform aims to provide a spot to design, deploy and monitor AI pods and data center workloads, Cisco said. Cisco anticipates initial trials of the new offering in Q4, followed by general availability.