Jump to Content
Security & Identity

The Future of Industries: How generative AI is altering the enterprise security landscape

February 12, 2024
https://storage.googleapis.com/gweb-cloudblog-publish/images/symantec-broadcom-gen-ai-security-intervie.max-2600x2600.png
Jeff Reed

VP, Product Management, Cloud Security

Symantec VP Adam Bromwich explains how AI is helping security teams detect and mitigate threats faster — including risks from other AIs.

Join us at Google Cloud Next

Early bird pricing available now through Jan 31st.

Register

Generative AI has been touted as the next productivity frontier. Analysts at McKinsey believe it can accelerate automating tasks by 10 years.

The potential for gen AI to have a positive impact on how we secure systems, networks, infrastructure, and organizations can’t be understated. It could help organizations identify and respond to threat alerts faster, freeing up security teams to focus on the further strengthening of defenses. At the same time, it can deliver broad business value including time savings and productivity gains.

Adam Bromwich, Vice President of Engineering and CTO for the Symantec Enterprise Division at Broadcom, believes generative AI could be the biggest security trend organizations have ever seen.

“The security industry is constantly evolving," Bromwich said. "It’s a moving target — and it’s what makes working in this space equally awesome and challenging. It’s not just that you have this active adversary that’s always evolving, but also that the technology is fundamentally changing, too. The latest change — gen AI — is going to absolutely reshape how enterprises protect their systems and data.”

Bromwich knows of what he speaks, having spent more than two decades working on enterprise security at Symantec, continuing through its 2019 acquisition by Broadcom. He recently sat down with Jeff Reed, Google Cloud’s vice president of Cloud Security, to discuss the changing nature of enterprise security and explore how gen AI is set to deliver wide-reaching business benefits in this space.

Jeff Reed: What do you think your customers will demand of enterprise security over the next three to five years?

Adam Bromwich: Across enterprises and industries, CEOs share a common and constant concern. That is, they need to know their data and information is safe from harm, and that the security systems they have in place will protect them. This won’t change any time soon. Broadcom’s customers will continue to expect us to stay one step ahead of new threats and attack vectors.

What has changed, though, is the threat landscape driven by mobile, cloud and now gen AI. In my many years in the industry, I’ve noticed that a shift occurs every couple of years as we rally around new technologies. Way back, there was behavior blocking. Then, there was sandboxing and advanced machine learning. They were all responding to the changing landscape and bringing with them new attack surfaces. And now, of course, by far the biggest change is gen AI.

https://storage.googleapis.com/gweb-cloudblog-publish/images/Untitled_ZMIJbQq.max-1700x1700.png

What role will technology play in shaping the future of this industry?

Bromwich: In the security industry, gen AI is going to be huge — both for us as defenders but also for attackers as well. Over the next few years, we’ll see a rise in the number and severity of attacks because gen AI will make it easier for attackers to automate a lot of the bad work they do.

On the flip side, more organizations will be implementing gen AI tools for security purposes. I think there’s going to be a huge opportunity around better and faster analytics, helping us predict attacks and become more preventative and proactive across the board.

What is a specific use case where gen AI could provide value to the industry?

Bromwich: We’re already using gen AI to make it easier for our customers to use our security products. Natural language search across the enterprise is proving incredibly effective. For example, it can search all the telemetry in your enterprise to understand what’s going on. This delivers invaluable time savings, enabling the security team to focus on solving problems and strengthening defenses.

Gen AI delivers huge time savings for developers, too. Developers can build new code through a simple chat interface to accelerate the deployment of cybersecurity programs. Ultimately, this new code can help protect organizations from harm.

Looking ahead, as security models get smarter, I think gen AI’s predictive analysis capabilities will soon prove their efficacy in weeding out weak points or vulnerabilities. For example, a gen AI tool could analyze a new environment — looking at things it has never seen before — and spot similarities with systems that have previously been compromised. Then, it could tell us, “These new things are very similar to those other things that led to an attack. To prevent an attack, you should disable them.”

And that’s just one security example. Across the board, the predictive capabilities of gen AI will find powerful applications in everything from fraud detection to insurance risk assessments. It’s going to make jobs easier for practitioners, for analysts, and for business productivity in general.

What gen AI use cases are you working on with Google?

Bromwich: We’ve just announced a project with Google Cloud that will give customers a significant technical edge for detecting, understanding, and remediating sophisticated cyber attacks. We’re embedding Google Cloud gen AI throughout Symantec’s portfolio, using natural language interfaces to generate more comprehensive and easy-to-understand threat analyses.

Previously, an analyst would have to pour over the raw data in a binary file and convert it into a consumable form. Now, all they need to do is have gen AI produce a human understandable description of the data. The time savings are huge — not only boosting the productivity of security operations center, or SOC, analysts and freeing their time to focus on combating bigger threats but ultimately helping enterprises respond to threats faster.

For this task, gen AI significantly improves analyst productivity, with answers available in seconds, not hours. It does 80% of the work, meaning our analysts only have to focus on the last 20% — which is pretty amazing. To take a Windows binary file and describe its threat capacity in a human readable format is a rare skill for analysts. Gen AI is a huge help here.

https://storage.googleapis.com/gweb-cloudblog-publish/original_images/image1_yUBSIA4.gif

Google Cloud's AI security tools can help generate code that saves time and addresses risks faster.

What started your AI journey with Google Cloud?

Bromwich: We’ve worked with Google Cloud for about four years now. Back then, we transitioned all of our products to Google Cloud because of the robustness of Google’s network at the time and the broad range of services and data center expertise that we needed.

With gen AI, we realized very quickly that building an AI model requires an enormous investment. Google was already building a security model, which we knew we could help train and then use ourselves. So, we’re essentially leveraging each other’s strengths.

Our core focus is on stopping threats. Sure, we love gen AI, and I wish we could have 1,000 people working on a gen-AI system of our own. But it would distract us from what we do best — and I’d much rather rely on Google Cloud, who is not only building a great gen AI system, but one with security at its core.

What is your boldest prediction about the future of security?

Bromwich:I think it has to do with how attacks will happen in the future. In the past, attackers targeted humans as the weakest link — starting with home users and moving more recently to those in large enterprises, stealing and using their credentials. We’re getting so much better at stopping these types of attacks.

But soon, attackers will be using gen AI to find and exploit vulnerabilities in infrastructure. And, let’s face it, as our systems become more complex in response to the needs of the modern workforce, more vulnerabilities are exposed. We’re unlikely to ever reach a point where there are zero vulnerabilities — instead, we need to focus our efforts on reducing the potential attack surface.

With this in mind, organizations should focus on building security into infrastructure from the start, and then using gen AI to identify vulnerabilities and fix them before the attackers get there. It’s turning into a big race.

https://storage.googleapis.com/gweb-cloudblog-publish/original_images/image3_30PfY2K.gif

What advice would you give to fellow industry executives looking to transform?

Firstly, keep at it. You need to keep transforming, even if it’s just to stay where you are today. You’ve also got to be willing to let go of the past and explore what’s possible, given this technological landscape.

Like a mountain with no top, transformation is an endless climb towards a peak of potential. As you climb, you need to keep applying the same passion — remembering that it’s not about the destination, it’s about the journey.


Opening image created with MidJourney, running on Google Cloud, using the prompt: a magazine-style illustration of a diverse team building a wall out of digital bricks that look like blue electricity. They're erecting the wall around an office building or city to protect it from hackers... can you make it look more secure, more like a construction than a finished thing, please ... make it look more like they're building it, and can you make the scene brighter, maybe more daytime looking, but otherwise very good.

Posted in