Artificial Intelligence (AI)
GTC 2025: AI, Security & The New Blueprint
From quantum leaps to AI factories, GTC 2025 proved one thing: the future runs on secure foundations.
Save to Folio
I hopped off the plane at LAX with my laptop bag and a regionally appropriate corporate vest and quickly made my way to San Jose, CA to kick off NVIDIA GTC 2025. If you feel artificial intelligence has already hit the peak of inflated expectations, I think you're wrong.
While the hype exists, the hype is real, and this was further punctuated by a doubling of attendees at the AI giant’s annual conference from 2024 to 2025, ballooning to 25,000 live attendees with an additional 300,000 virtual attendees tuning in.
The sessions lent insight into a future where AI is inextricably linked to every facet of our lives—from manufacturing to climate crisis solutions to healthcare to drug discovery.
Whether the conversation started with AI factories, agentic reasoning, or quantum computing breakthroughs, they eventually led to where security will be integrated into each layer of this next generation in technology.
Everyone has a blueprint, and if you don’t — you’re late
NVIDIA's "blueprints" concept reaches beyond technical documentation and has become the common language among technology builders to compare requirements. These comprehensive frameworks map the use cases, outcomes, and technologies needed to build and scale AI solutions regardless of industry.
Examples such as Omniverse (digital twins) and AI-Q (agentic systems) streamline AI deployment through standardization and process consistency. Relevant for CISOs and CTOs alike, these blueprints make it easy to identify where cybersecurity measures fit in at the initial design phase, significantly reducing the potential for vulnerabilities and improving security resilience.
AI Factories: what are they?
A new collaboration between Dell and NVIDIA has resulted in the AI Factory initiative — a combination of computing infrastructure, data management, software and tools, security, and reference architecture (blueprints). The goal? Enable secure AI scale and rapid time to deployment. Essentially, it is making it possible for organizations of any size to efficiently scale AI solutions custom to their unique business requirements. Given the AI factories' broad reach into the enterprise digital infrastructure, proactive cybersecurity monitoring and insight into software vulnerabilities, data movement, and misconfigurations are critical to its success.
Quantum has arrived
GTC 2025 brought quantum computing from theoretical discussion to practical implementation as NVIDIA announced its Accelerated Quantum Research Center and use of the NVIDIA CUDA-Q quantum development platform to enable researchers to develop new hybrid quantum-classical computing algorithms and applications.
These advances — while exciting — come with some highly relevant cybersecurity implications, primarily concerning encryption vulnerabilities. Forward-looking organizations will be thinking of quantum-resilient encryption implementation as a future-proofing exercise to protect sensitive data and IP.
Agentic is the new Generative
NVIDIA's new Llama Nemotron models are a serious leap in agentic and reasoning AI. The implications are broad, but I’m particularly interested in its incoming role as a proactive cybersecurity tool to analyze high volume and complex data sets through multi-step, logical “thinking.”
Proactive security starts here
After the launch of ChatGPT at the tail end of 2022, organizations started rapidly — or tentatively — adopting new productivity applications. Some even started building their own. For the latter group my recommendation was consistent, “the most powerful apps are the most well protected apps.” This continues to ring true as the innovation emerging from GTC 2025 will accelerate industries and expand the “AI footprint” in our lives with more prolific and powerful technologies.
The innovations showcased at NVIDIA's GTC 2025 redefine the relationship between AI advancement and security excellence. Throughout the conference, Trend Micro leaders — Rachel Jin (CEPO), Fernando Cardoso (VP, Product Management), Patrick Lu (Lead AI Architect) — reinforced this vision of integrated security as a business accelerator — that secure by design isn’t a technical requirement, it’s a foundation for innovation at speed and scale.
The future belongs to organizations and AI innovators who identify security early as an enabler and integrate proactive security frameworks directly into their AI strategy. With this approach, organizations can accelerate innovation, limit risk early, and ultimately drive more powerful technology outcomes and productivity gains.
Learn more about the Trend Vision One Blueprint for AI Security and Trend Cybertron: https://newsroom.trendmicro.com/2025-03-19-Trend-Micro-to-Open-source-AI-Model-and-Agent-to-Drive-the-Future-of-Agentic-Cybersecurity