Nvidia CEO: Cybersecurity Is AI’s Next Big Market portrait

Nvidia CEO Jensen Huang Says Cybersecurity Could Be AI’s Next Big Application

Artificial intelligence has already transformed the way software is written. AI coding assistants can generate code, explain programming errors, suggest fixes and help developers complete tasks faster.

But what happens when AI moves beyond writing software and starts continuously checking whether that software is safe?

NVIDIA CEO Jensen Huang believes cybersecurity could become one of AI’s next major application areas. Speaking at the Goldman Sachs Communacopia + Technology Conference in San Francisco on September 10, 2026, Huang said cybersecurity would likely be the next major use case for AI after coding.

His prediction points towards a major change in the cybersecurity industry: moving from security tools that mainly respond to alerts and human instructions towards AI systems that can continuously search for weaknesses, investigate suspicious activity and support defensive operations.

From AI coding to AI cybersecurity

Huang connected the growth of AI coding tools with the growing need for cybersecurity.

AI systems are becoming increasingly capable of writing and modifying software. But software can contain vulnerabilities, and those vulnerabilities can be exploited by attackers.

As the amount of code being generated and modified increases, the need to find bugs and fix them also grows.

Huang described bug finding as a natural extension of coding. He then connected bug finding to cybersecurity, which he described as a much larger market.

The logic is straightforward: if AI helps create software faster, organisations also need better ways to check that software, identify weaknesses and protect the systems running it.

This creates demand for both software testing and cybersecurity services.

Why continuous AI security matters

Traditional cybersecurity tools often rely on predefined rules, alerts and security teams investigating suspicious activity. Human experts remain essential, but they cannot manually inspect every system, event and network connection at every moment.

AI could help close some of these gaps.

An AI-powered security system could potentially:

- Scan software and infrastructure for known vulnerabilities.

- Identify unusual network behaviour.

- Prioritise security alerts based on potential risk.

- Help security teams investigate suspicious activity.

- Recommend patches and defensive actions.

- Monitor systems continuously, including outside normal working hours.

The important difference is that these systems could operate continuously instead of waiting for a person to submit a prompt.

However, continuous operation does not automatically mean reliable protection. AI systems can make mistakes, overlook threats or generate false alarms. Organisations still need access controls, testing, monitoring and human approval for high-impact actions.

The two sides of AI: attacks and defence

The cybersecurity opportunity has two sides.

On the offensive side, malicious actors can potentially use AI to speed up vulnerability discovery, automate parts of cyber operations and make attacks more scalable.

On the defensive side, security teams can use similar technical capabilities to identify weaknesses before attackers exploit them.

This creates an important security challenge: defenders need to use AI effectively without allowing attackers to gain a greater advantage.

The same general capability that helps identify a weakness can be useful to someone trying to exploit it. The outcome depends on who controls the system, what permissions it has and what safeguards are in place.

That is why AI cybersecurity is not simply about installing a chatbot inside a security department. It requires carefully designed systems, restricted permissions, reliable testing and clear accountability.

NVIDIA is already moving into this market

Huang's comments are not just a prediction about a distant opportunity. NVIDIA is already working with companies in the cybersecurity ecosystem.

At CrowdStrike's Fal.Con 2026 event, NVIDIA and CrowdStrike discussed SafeMind, an agentic cybersecurity system combining CrowdStrike technology with NVIDIA Nemotron models. NVIDIA described the system as using offensive and defensive AI capabilities in a continuous process of testing and improving cyber defence.

Huang also discussed partnerships involving CrowdStrike, Cisco and Palantir at the Goldman Sachs conference.

These partnerships show how NVIDIA's role could extend beyond selling GPUs. Its AI models and computing platforms can become part of the infrastructure used by cybersecurity companies to develop and operate AI-powered security products.

That does not mean every product is already fully autonomous or capable of preventing every attack. It shows where companies are investing and what kind of security systems they are trying to build.

Why businesses may pay for AI cybersecurity

For businesses, cybersecurity is not an optional technology expense. A successful attack can disrupt operations, expose customer information, damage reputation and create financial losses.

Companies therefore have a strong reason to invest in better protection.

AI could make security operations more efficient by helping teams process large volumes of information, identify important alerts and investigate incidents faster.

For cybersecurity providers, continuous AI services could also create recurring business opportunities. Organisations need protection every day, not just when a security incident occurs.

This is one reason Huang sees cybersecurity as a major commercial opportunity for AI companies and established security vendors.

But businesses will not adopt these systems simply because they use AI. They will need evidence that the products reduce risk, improve response times and justify their cost.

What does this mean for India?

India is a significant market to watch because businesses, banks, hospitals, government services and digital platforms increasingly depend on connected systems.

As more organisations move operations online, the consequences of weak cybersecurity can become more serious.

AI-powered security tools could help Indian organisations monitor systems, identify suspicious activity and support smaller security teams dealing with growing workloads.

There may also be opportunities for Indian cybersecurity startups and software professionals to build products around threat detection, vulnerability management, security automation and AI-assisted incident response.

For students, this creates an interesting career direction. Learning AI alone may not be enough. Combining programming, networking, operating systems, cloud computing and cybersecurity fundamentals could open up more specialised opportunities.

The challenges AI security companies must solve

Despite the opportunity, AI cybersecurity has serious limitations.

False alarms: An AI system may incorrectly identify normal activity as an attack, wasting the time of security teams.

Missed vulnerabilities: AI may fail to recognise a weakness or misunderstand the context in which software is running.

Data privacy: Security systems may need access to sensitive logs, network information and business data. Companies must control how that information is stored and processed.

Autonomous actions: Giving an AI system permission to block users, change configurations or modify infrastructure can create new risks if it makes an incorrect decision.

Adversarial manipulation: Attackers may try to trick AI security systems or exploit the systems themselves.

For these reasons, AI should strengthen security teams rather than become an unchecked replacement for them. Human oversight, restricted permissions, independent testing and clear incident procedures will remain important.

The bigger picture

Jensen Huang's prediction reflects a wider transition in artificial intelligence.

The first major wave brought AI assistants into everyday work. Coding became one of the clearest examples of how AI could improve productivity. The next wave may involve AI systems that perform ongoing work in the background, including monitoring, testing and defending complex digital infrastructure.

Cybersecurity is a natural candidate because threats do not follow office hours and security teams must constantly respond to changing conditions.

But the opportunity comes with a difficult reality: AI can help defenders, and it can also help attackers.

The companies that succeed will not simply be those with the most powerful models. They will be the ones that can make AI systems dependable, secure, measurable and accountable.

Huang sees cybersecurity as a major next market for AI. Whether it becomes one of the industry's biggest applications will depend on how well these systems perform in the real world.