Singapore's rush to adopt AI is increasing cybersecurity risks

The rise of AI is unassailable, driven by the promise of competitive advantage and enhanced business performance.

We spoke to business leaders and CTOs in Singapore, and 82% were wary that the fast adoption of generative AI (Gen AI) will have long-term repercussions on their IT infrastructure. At the same time, 80% of IT teams are under increasing pressure roll out Gen AI within their organisation.

Given the rapid evolution of Gen AI, it's critical that organisations are fully prepared for its adoption. While there are many opportunities created by Gen AI, there are risks. Most notably, security and data sovereignty - it's crucial to understand and address these risks to safeguard your organisation.

The Gen AI risks

Gen AI is based on models that are trained using vast amounts of data and processing power. This data is constantly being transferred from the user's network environment and the AI tools cloud network which is usually a public cloud environment. This creates two risks.

First, the constant stream of data transiting into and out of an organisation's network environment to the public cloud. If your organisation is connecting over the public internet, it is not a question of whether a breach will happen but when. More than half of IT leaders in Singapore (58%) say Gen AI is likely to put their network at risk of cyber-attacks compared to 71% of IT leaders globally. This suggests Singaporean organisations are either less reliant on the public internet than in other regions or they are less aware of the risks.

Unmanaged use of Gen AI in enterprise applications can also introduce risks such as sensitive data exposure or inadvertent disclosures. The other notable risk associated with Gen AI is linked to network resilience, associated with large amounts of data being transferred across network environments. Our research found an alarming 81% of IT leaders in Singapore believe their network infrastructure doesn't have the capacity to embrace Gen AI to its full potential. Without the appropriate infrastructure in place, Gen AI has the potential to increase latency and congestion across a network, creating an inadvertent, internal denial of service attack.

Network-as-a-Service (NaaS) to mitigate risks

With Gen AI and other cloud-based tools rising in popularity, networks are handling larger volumes than ever before. A NaaS platform helps organisations to adapt their connectivity to accommodate the unpredictable traffic patterns generated by AI workloads.

NaaS provides the flexibility to scale network resources based on demand, ensuring that businesses can efficiently handle spikes in AI-driven traffic without compromising performance. This adaptability is crucial for AI applications that require consistent, real-time data processing, as NaaS optimises connectivity to support high-speed, low-latency performance.

In addition to performance benefits, NaaS can enhance security by removing traffic from the public internet, using private network infrastructure. This secure data transfer protects sensitive information, preventing unauthorised access and potential breaches. Organisations can also prioritise traffic - granting higher priority for sensitive dataflows or mission critical processes. 

Reduce costs, as well as boost efficiency.

The shift towards the cloud has required a mindset shift for financial operations. Gone are the days of significant up-front costs to establish your own infrastructure, instead organisations pay for what they use on the cloud.

Not only does AI require an increase in capacity, but it also transfers large amounts of data to and from the cloud, increasing egress charges. NaaS helps optimise data transfer and reduce the need for traditional network infrastructure with can reduce egress fees by as much as 50%.

A NaaS platform also enhances operational efficiency by speeding up data transfers and simplifying service provisioning. This is especially beneficial for AI applications that demand rapid data processing and minimal latency. It can also improve cloud access by providing seamless connectivity between various cloud platforms, regions, and on-premises data centres. This capability supports robust hybrid and multicloud strategies, leading to more flexible and efficient utilisation of cloud resources.

Final thoughts

Gen AI offers organisations vast potential. It can streamline processes and operations, create new business models and improve people's quality of work. But it is not something that should be deployed ad-hoc.

As the old saying goes, failure to prepare is preparing to fail. Organisations need to carefully consider their network infrastructure and capacity before deploying AI, otherwise they increase their exposure to security and performance issues. Network-as-a-Service could be the tool that helps organisations balance the networking demands of Gen AI while mitigating risks around security, performance and costs.