Introduction
Artificial intelligence (AI) dominates cybersecurity headlines in 2025. Some vendors promise AI will revolutionise defence and end cybercrime altogether. Others paint a darker picture: AI-powered hackers outsmarting every system.
For Irish enterprises, the truth lies somewhere in the middle. AI is already changing how attacks are launched and how defenders respond — but separating hype from reality is critical for CEOs and boards making investment decisions.
This blog examines how AI is being used in cybersecurity today, where it genuinely adds value, how criminals are exploiting it, and what Irish business leaders must know before betting their strategy on it.
1. The Hype: AI as a Silver Bullet
Marketing hype around AI in cybersecurity is everywhere. Vendors claim their platforms can:
- Predict and prevent attacks before they happen.
- Eliminate the need for human analysts.
- Block every phishing attempt automatically.
This narrative has fuelled enormous spending. Analysts project the global AI-in-cybersecurity market to reach $46 billion by 2027.
But here’s the problem: AI isn’t magic. It’s a tool, and like any tool, its value depends on how it’s designed and deployed. Without the right context, training data, and human oversight, AI can create just as many problems as it solves.
💬 “Any CEO who thinks they can buy one AI product and switch off their people is taking a dangerous gamble,” warns Gary Melvin, CTO of Nostra. “AI augments humans, it doesn’t replace them.”
2. The Reality: Where AI Delivers Real Value
When implemented responsibly, AI can significantly improve cybersecurity defences. Areas where AI is proving its worth include:
- Threat detection and response
Analysing huge volumes of network traffic to spot anomalies faster than human teams.
- Phishing detection
Reviewing millions of emails for signs of fraud, beyond what filters can catch.
- Automated patching
Identifying and prioritising known vulnerabilities across complex systems.
- Incident triage
Freeing human analysts from routine alerts so they can focus on complex threats.
Irish firms already using Microsoft Sentinel’s AI-driven analytics have seen detection and response times cut from hours to minutes. But crucially, AI isn’t working alone — it’s being guided and verified by human security teams.
💬 Darren Mooney, Head of Security at Nostra, explains: “AI can handle the haystack, but only people can recognise the needle. If you rely entirely on AI, you’ll either drown in false alarms or miss the real threat.”
3. AI in the Wrong Hands: How Hackers Exploit It
Unfortunately, cybercriminals are just as quick to weaponise AI. In 2025, we’ve seen:
- Deepfake CEO scams: AI-generated video and audio messages convincing finance teams to transfer funds.
- AI-driven phishing: Personalised emails written in flawless Irish-English, making them far harder to spot.
- Automated vulnerability scanning: AI bots probing Irish enterprise systems around the clock.
In late 2024, an Irish financial services firm narrowly avoided a €1.2 million fraud when a finance executive received what appeared to be a video call from their CFO. Only a manual verification step stopped the transfer.
💬 Reg Stallard, Cyber Risk Consultant, says: “We’ve seen AI-generated emails with perfect tone, local references, and no errors. They’re almost indistinguishable from the real thing.”
👉 Key Point: AI threats aren’t theoretical. They’re here, and they’re targeting Irish enterprises right now.
4. Cutting Through Vendor Hype
The challenge for CEOs is separating real capability from marketing spin. Red flags when evaluating AI cybersecurity products include:
- Guarantees of 100% protection (no system can promise this).
- “Black box” algorithms with no transparency on how decisions are made.
- Solutions trained only on overseas data that may not recognise European or Irish-specific attack patterns.
💬 “When we assess AI tools, the first questions we ask are: what data was this trained on, and how do you test for bias or blind spots? If vendors can’t answer, that’s a deal-breaker,” notes Darren Mooney.
5. Responsible Adoption of AI in Security
Irish enterprises need to take a measured, responsible approach to AI adoption:
- Start with your risks: Identify where AI will make the most impact — not just where it looks impressive.
- Keep humans in the loop: Use AI to support analysts, not replace them.
- Audit algorithms regularly: Ensure they perform as expected and comply with GDPR.
- Integrate into a wider strategy: AI is one layer of defence, not the whole wall.
At Nostra, we pair AI-driven monitoring tools with 24/7 human response teams. This hybrid approach ensures speed without sacrificing judgement.
6. What Boards Should Be Asking in 2025
Boards should treat AI in cybersecurity the same way they treat financial or compliance investments — with governance and accountability.
Questions directors should be asking include:
- How is AI being used in our cyber defences today?
- What are the risks if attackers use AI against us?
- How do we verify that AI tools are effective and compliant?
- Are our staff trained to spot AI-generated fraud attempts?
💬 CTO Gary Melvin sums it up: “Boards need to stop seeing AI as a black box. It’s a tool, and its impact depends on how it’s governed.”
7. The Future: AI in 2026 and Beyond
Looking ahead, AI will bring both opportunity and risk. On the positive side, it will:
- Make threat prediction more accurate.
- Reduce false positives in monitoring systems.
- Correlate cross-industry threat intelligence faster.
But we must also prepare for AI-powered malware that can rewrite itself, and the long-term risk that quantum computing combined with AI could break encryption standards.
👉 For Irish enterprises, the message is clear: AI must be monitored, governed, and continually reassessed.
Conclusion
AI is transforming the cybersecurity landscape on both sides of the battlefield. For defenders, it offers faster detection and sharper insights. For attackers, it provides new ways to deceive, automate, and scale.
The winners in 2025 will be Irish enterprises that cut through the hype, adopt AI responsibly, and keep people firmly in the loop.
👉 Book your free cybersecurity assessment with Nostra to see how ready your organisation is for AI-powered threats.