Table of Contents
Speaking at the recent Cisco AI Summit, Jeetu Patel, the company’s president and chief product officer, outlined what he described as three major constraints holding back AI
In sum – what to know:
Infrastructure limits slow AI scale – Power, compute and network bandwidth shortages remain the biggest constraint.
Trust becomes prerequisite for adoption – Cisco says security and confidence in AI systems are now essential conditions for enterprise deployment.
Data shortages reshape model training – As public data runs out, synthetic and machine-generated data are becoming central to future AI development.
Cisco executives said infrastructure limits, trust concerns and data shortages remain the main barriers to large-scale AI adoption, even as enterprise deployments accelerate.
Speaking at the recent Cisco AI Summit, Jeetu Patel, the company’s president and chief product officer, outlined what he described as three major constraints holding back AI: infrastructure capacity, trust in AI systems and the availability of training data.
“The first one is that we have an infrastructure constraint. We just don’t have enough power, compute and network bandwidth. So that’s going to be an area that we are spending billions on at Cisco. And I think the industry is spending trillions in making sure that we can… go out and fulfill the needs of AI,” Patel said. “We’re working very hard to make sure that we can build out the critical infrastructure for AI.”
He pointed to Cisco’s P200 chip and the Cisco 8223 routing system, announced last October, as core technologies for building AI clusters across multiple data centers.
“The P200 chip was for the scale out, because what’s happening now is these models are getting bigger where they don’t just fit within a single data center. You don’t have enough power to just pull into a single data center,” Patel said. “Now you need to have data centers that might be hundreds of kilometers apart, that operate like an ultra-cluster that are coherent. And so that requires a completely different chip architecture to make sure that you have capabilities like deep buffering and so on and so forth… You need to make sure that these data centers can be scaled across physical boundaries.”
He also pointed out that the industry is reaching the physical limits of copper and optics, with coherent optics rising in importance as data center infrastructure is built out. “That’s an area that you’re starting to see a tremendous amount of progress being made,” Patel said.
The second constraint is trust, he added. “We currently need to make sure that these systems are trusted by the people that are using them, because if you don’t trust these systems, you’ll never use them,” Patel said.
“This is the first time that security is actually becoming a prerequisite for adoption. In the past, you always ask the question whether you want to be secure, or you want to be productive. And those were kind of needs that offset each other,” Patel said. “We need to make sure that we trust not just using AI for cyber defense, but we trust AI itself.”
The third constraint is a growing data gap. AI models have largely relied on publicly available, human-generated data, but “we’re running out,” Patel said. Synthetic data is becoming increasingly important as models scale, while machine-generated data is growing rapidly as autonomous agents produce information continuously.
He also noted the growing role of AI in software development inside the company. “It seemed like a far-fetched goal at the beginning of 2025, but now 70% of all AI products now being developed at Cisco are using code that’s generated by AI. I would say that within the year 2026, we will have at least close to a half a dozen products that’ll have 100% of the code written by AI rather than written by humans,” Patel said. “Humans will still have a very valuable role to play because they’re going to make sure that they’re writing specs and they’re making sure that they’re actually going out and reviewing the code. But the bottleneck is no longer going to be around the writing of the code activity. The bottleneck is going to be around the reading and reviewing of the code activity.”
Cisco CEO Chuck Robbins said 2026 will mark a turning point for enterprise AI adoption, driven by the emergence of agentic applications and a greater focus on trust and security.
“There are lots of questions and discussions about what does it mean to your enterprise infrastructure? What does it mean to your security posture? What does it mean to application development cycles?” Robbins said.
“One thing that bothers us is trust, where there’s trust in what’s going to happen to your data, trust in the models, trust in your infrastructure, trust in the agents, trust in the partners that you’re working with – those are important issues that the industry needs to continue to address with AI going forward,” Robbins said.
