Combination of humans and AI paramount to dealing with agentic AI cyberthreats

“We are very clear that the future is a combination of humans plus AI. That's the magic bullet, because I'll tell you, humans have contextual knowledge, which AI systems, at least at this point, don't have,” says Rashish Pandey, Vice President, Marketing & Communications, APAC at Fortinet.

Looking at the 2026 threat landscape, Fortinet predicts AI threats to become increasingly more sophisticated and harder for organizations to detect and deal with. With cybercriminals now even leveraging agentic AI capabilities to create and launch more sophisticated threats, Fortinet also believes the world could see the first ever purpose-built cybercrime agent in 2026.

“We have not seen full agentic cybercrimes happen yet. We forecast that 2026 might be the year we'll see the first sort of agentic overgrowth. Humans will always be in the mix on both sides of the house. It will never be a fully automated, completely hands-off operation. We will see a lot of AI and automation coming in on the defender side because today it's very siloed. Remember, threat actors have to get in only one time while the defenders have to defend all the time,” commented Rashish Pandey, Vice President, Marketing & Communications, APAC at Fortinet during a media briefing in Kuala Lumpur.

Pandey explained that organizations will need to have a cybersecurity posture which is a lot more robust and broader.

“What has happened is historically, the way we have procured cyberinfrastructure is very piecemeal. For example, if I need email security, I go buy an email security system. If I need a web security system, I go buy a web security system and such. What happens is customers end up having 20-30 different cyber systems. And it's very hard to integrate a common playbook or integrate a common detecting response framework when there is such a diverse landscape,” said Pandey.

At the same time, Pandey pointed out that customers are now platforming, which means they are trying to reduce the number of platforms in their infrastructure. They are also trying to automate cybersecurity.

“The next level is to AI power that automation so that the humans, while still in the loop at critical decision points, can get a lot of the gradual work and a lot of the high-volume work automated at much faster machine speeds,” added Pandey.

Focus on velocity, scale and adaptability

According to Pandey, in 2026, organizations will need to focus on velocity, scale and adaptability as they look to be better prepared to deal with cyberthreats. As automation now accelerates almost every form of attack, cyber defenders need to unify threat intelligence, exposure management and automotive response into one adaptive system.

Increasing dependence on AI will only lead to increased need to manage identities as well. Identity management for both human and machines needs to be the foundation of trust and the primary control surface for organizations. This is imperative for organizations because humans and AI need to operate together to ensure they are well prepared to withstand the emerging era of AI-driven cybercrime.

“It’s not just about AI for AI cases; it's about the capabilities it delivers, both on the offensive side as well as on the defensive side. If you could do it without using AI, it's possible, but we don't see it. The offenders will continue to use AI. We need to use AI to stay at the same speed, if not ahead of the attackers,” explained Pandey.

When asked how Fortinet is working with governments in the region to be better prepared to deal with these threats, Pandey believes that cyber threats are global and not just local.

“A lot of the groups targeting a country like Malaysia are actually globally engaged. So, the response needs to come through a globally coordinated attack, which includes working with local law enforcement as well. So, when an attack happens and it's detected locally through one of our customers or through our systems, we actually work with the customer to notify local law enforcement about the issue and put in support and things are actually made accordingly. We also share threat intelligence with the relevant government agencies because they are on the same page,” said Pandey.

Pandey also mentioned that a lot of countries are seeing new cyber laws coming in, especially targeting the critical infrastructure side, so that there is a standardized way of disclosure of breaches and responses that customers can do.

Meanwhile, Kevin Wong, Country Manager, Fortinet Malaysia pointed out that a lot of SMEs in Malaysia may not have a very large cybersecurity body base.

“So, we work with technology partners, managed service partners, and telco providers to actually provide a whole suite of solutions to help them manage the cyber security threat landscape for SME customers,” said Wong.

The regional approach

As ASEAN is a very dynamic environment, Pandey believes that almost every country in ASEAN is rapidly digitizing and is focusing on manufacturing, tourism, enhancing connectivity, and such, which is a fantastic situation from an economic point of view.

“But remember, every time you connect and every time you integrate, you also create cybersecurity possibilities of attackers getting in. So the very important thing is, as countries digitize, we need to make sure that wherever we are focusing on digitization, whether it's manufacturing, supply chain, or rebalancing of the whole manufacturing footprint across the globe, we need to build cyber security from the start and not as an afterthought,” explained Pandey.

Interestingly, Pandey also pointed out that from a geopolitical perspective, any form of conflict from this point onwards will happen on the cyber field.

“It will not be a bullet; it will not be a missile. Any geopolitical threshold will start with a cyberattack or a cyber offensive. We have seen it in the conflicts happening across the region; it's only going to accelerate. So, as we build national defenses and as we build collaborations across the region, it's really important to remember that cyber is the first battleground in any conflict,” added Pandey.

As such, Pandey firmly believes that the focus in 2026 will not just be on AI but on how fast organizations respond with their entire footprint. This means the focus needs to be on agility and velocity as a design principle.

“We are very clear that the future is a combination of humans plus AI. That's the magic bullet, because I'll tell you, humans have contextual knowledge, which AI systems, at least at this point, don't have. So, the future is human teams plus AI teams working as one unified organization,” concluded Pandey.