In 2025, an invisible conflict is in progress. AI-powered malware is constantly attempting to destroy your company, while AI algorithms are there to safeguard it. Every second, both systems are learning, evolving, and becoming (‘generatively’ stronger) creative. Or disruptive? The real question isn’t if your defences will hold for the day, but whether your AI can out-think the one trying to tear your digital fortress apart.
As AI evolves, so too does the battle it fuels — between Good AI, which defends, and Bad AI, which disrupts. The question is whether your AI is agile enough to outpace the Bad AI trying to undermine your security — constantly adapting, learning, and evolving with an intelligence of its own.
Indeed, Singapore’s position as a global and regional hub makes it a prime target for bad actors, many of whom are AI-enhanced. To maintain a strong front in the digital battlefield, Singapore must take cyber resilience seriously.
A catalyst for defense and a tool for threat
No longer just a buzzword in boardroom conversations, AI has advanced significantly in recent years. Although it has the potential to change people’s lives, it also creates a gap for far more serious problems.
AI is now central to modern cybersecurity. AI has become essential due to its capacity to handle massive amounts of data, identify and even anticipate problems, and react quickly. Malicious actors, however, have used that same power to significantly increase the sophistication and frequency of cyberattacks.
AI-powered malware has evolved into a potent danger that can imitate normal behaviors, avoid detection, and constantly adjust to defenses. Malware is typically identifiable by its patterns, which allow traditional antivirus software to detect it. However, with AI, hackers can produce malware that dynamically modifies its behaviour and code to avoid detection, imitating trustworthy operations and changing its structure with every infection.
According to research by the World Economic Forum, cybercrime costs are expected to reach $10.5 trillion annually by 2025. In Singapore, the total amount lost in 2024 was at least $1.1 billion. And it doesn’t end there. These attacks will increase in frequency and difficulty as the AI era progresses. However, AI is not solely about the adversarial. Trust, transparency, and human alignment are the main goals of good AI. It is intended to preserve privacy, ethics, and security while also evolving responsibly. In stark contrast, bad AI exploits its power — hiding behind layers of opacity, bias, and harmful intentions. For instance, the AI-driven malware Emotet began as a banking Trojan but has since developed into a very serious and advanced threat. Emotet’s ability to change its behaviour through artificial intelligence makes it extremely challenging for conventional security measures to identify and neutralize. There has never been a more pressing need to transition from reactive defence to proactive, AI-driven protection.
Building a resilient defense
As we embrace AI’s role in cybersecurity, there is an imperative: to stay one step ahead, we must think beyond reactive measures. To safeguard against this new wave of attacks, enterprises must adopt a proactive approach.
1. AI-powered threat detection and response
Predictive threat intelligence platforms can anticipate attack vectors by analyzing massive datasets to identify emerging attack patterns and block threats autonomously, minimizing human intervention.
These platforms, offered by leading players in the commtech industry, continuously adapt to new threats, providing early detection and real-time protection.
2. Zero Trust with AI-enhanced access control
Zero Trust is crucial in the modern world, but when paired with AI, it will become much more powerful in the future. Real-time risk evaluations will be offered by AI-powered Identity and Access Management (IAM), depending on variables including user behavior, location, and device health. By ensuring that authorized individuals can access sensitive data, this ongoing analysis will significantly reduce the risk of both external and insider attacks.
3. Self-healing networks
AI will enable self-healing networks that can recognize and eliminate threats independently. AI-powered self-healing systems can automatically identify security breaches, isolate compromised systems, and then restore them to a secure state without the need for human intervention, guaranteeing that the company will continue to function even in the event of an attack.
4. Blockchain for data integrity
As AI continues to permeate every part of our digital life, ensuring data integrity has never been more crucial. Together, blockchain and AI provide a second layer of security that protects transactions in real time, ensures data authenticity, and eliminates tampering.
5. Collaborative threat intelligence
Futuristic AI-powered platforms will enable organizations to share threat intelligence globally. This will allow enterprises to collectively stay ahead of increasingly sophisticated attacks, strengthening the digital fabric of industries across the globe.
The future isn’t just about deploying AI for defence but understanding how it works, how it evolves, its capabilities, limitations, and how it can fail. Equip and train your teams to understand AI, update your strategies, and keep your defences as dynamic as the threats you’re up against.
As we move forward into an AI-driven future, businesses that will thrive are those that not only use AI to secure their digital assets but also ensure it remains aligned with the greater good. In this ongoing battle, it’s not about having AI. It’s about having the smarter AI.
Bhaskar Gorti is Executive Vice President, Cloud & Cybersecurity Services, Tata Communications.
TNGlobal INSIDER publishes contributions relevant to entrepreneurship and innovation. You may submit your own original or published contributions subject to editorial discretion.
Featured image: Andres Siimon on Unsplash
Sovereign AI: The new strategic imperative for governments and enterprises