The exploration of AI beyond mere hype: Guidelines for security professionals to create AI agents with significant impact
In today's digital landscape, where cyber threats are constantly evolving, the need for efficient and secure cybersecurity solutions has become paramount. One such solution is Tines, a Dublin-based startup that specialises in AI-powered workflows. Founded in 2018 by Thomas "TK" Kinsella and Eoin Hinchy, Tines became a unicorn in February 2025.
Secure Architectures for AI Agents
Granular access controls are essential in agent secure architectures, following the principle of least privilege. This means that each agent is given only the necessary permissions to perform its tasks, minimising the potential for security breaches.
Security teams should prioritise agents they can trust and verify, ensuring that every agent action is auditable and revocable. Zero data exfiltration is a priority in these architectures, with the AI agents built by Tines running inside a secure infrastructure and not exfiltrating data.
The Role of AI Agents in Cybersecurity
AI agents can automate repetitive tasks in cybersecurity, such as alert enrichment, threat scoring, and device isolation. These agents can help teams do more with less, respond faster, and stay focused on high-impact work.
Moreover, AI agents can eliminate alert fatigue, reduce burnout and turnover, and strengthen security postures through offloading repetitive, mindless muck work. In modern SOCs, teams can use agents for repetitive tasks like malware submissions to VirusTotal, while using copilots for guidance during threat hunting.
Balancing Automation and Security
While AI agents offer numerous benefits, over-investment in AI should be guarded against. It's crucial to ensure that automation is used with purpose and not as a distraction. Security leaders should build frameworks, track impact, and insist on secure design to reclaim their team's time, confidence, and capacity to lead.
The level of autonomy for AI should be matched to the task at hand. Deterministic automation is suitable for rule-based actions, copilots provide human assistance, and agents are designed for independent decision-making.
Corporate stakeholders are also interested in understanding the risk calculus of their technology stacks, seeking answers to questions like "Are we a target?". By implementing secure and efficient solutions like Tines, organisations can better protect themselves and make informed decisions about their cybersecurity strategies.
