AI in Cybersecurity: Boon or Risk for SMEs?

Artificial intelligence (AI) is becoming a double-edged sword in cybersecurity. While it helps defenders automate responses and detect anomalies, it also empowers attackers to craft more sophisticated threats. For New Zealand SMEs navigating resource constraints, understanding how to use AI securely is key to staying ahead.

🧠 What AI Brings to Cyber Defence

AI-based tools, particularly those using machine learning, analyse network traffic and user behaviour at scale. They flag deviations—like unusual login times or large data transfers—and generate alerts faster than human analysts. For SMEs without a dedicated security team, this capability is transformative.

🤖 Attackers Use AI Too

Unfortunately, cybercriminals also use AI. They automate phishing, mimic human communication patterns, and dynamically adapt malware to avoid detection. The result is a surge in highly convincing email attacks, deepfake videos for fraud, and polymorphic ransomware variants that mutate with every attempt.

🔍 AI’s Role in Threat Detection and SOC Tools

Modern Security Information and Event Management (SIEM) tools integrate AI to prioritize alerts, identify false positives, and correlate threats across endpoints and cloud services. SMEs leveraging these tools gain early visibility into threats—but only if they tune and maintain the system regularly.

⚖️ The Risk of Over-Reliance

While AI offers power, it doesn’t replace fundamentals. Over-relying on AI tools without human oversight creates blind spots. SMEs must maintain layered defences, enforce access policies, patch regularly, and train staff. AI is an enhancer, not a silver bullet.

📈 AI-Driven Compliance and Reporting

Regulatory obligations are evolving fast. AI now assists in mapping data flows, identifying risk areas for ISO/NIST compliance, and generating audit trails. For resource-stretched SMEs, this helps demonstrate governance to clients, regulators, or insurers.

🔐 Choosing AI Tools Wisely

Before deploying AI-based security tools, SMEs must ask:

• What data will the tool access and store?
• Is the AI model explainable and auditable?
• Does it integrate with our existing environment?
• Can we sustain the operational cost and training?

Many vendors overhype their AI features—validation through pilots or proof of concept is critical.

🔄 AI in Incident Response

AI helps speed up triage and containment during breaches. From identifying affected assets to recommending isolation strategies, AI reduces the time to act. However, SMEs must still have human-led playbooks in place to review and execute responses.

🔒 Local Context: NZ Privacy & Ethics

In New Zealand, AI use in security must comply with the Privacy Act 2020. SMEs must ensure tools don’t introduce unnecessary surveillance or privacy risks, especially when handling customer data. Transparent use of AI builds trust, while opaque automation erodes it.

💡 Bottom Line

AI in cybersecurity offers real benefits for SMEs—but only with strategic implementation, proper controls, and realistic expectations. It’s not about chasing the latest AI trend; it’s about making your existing security posture more intelligent, responsive, and resilient.

Eduardo Wnorowski is a Technologist and Director at Virtus Group Ltd.
With over 29 years of experience in IT and consulting, he brings deep expertise in networking, security, infrastructure, and transformation.
Eduardo helps New Zealand businesses navigate change with clarity, security, and trust.
🔗 Connect on LinkedIn