Combining AI with specialist judgement and clear accountability can turn cybersecurity into a business advantage, Amusnet’s Chief Information Security Officer Stoyan Popov explained at Digital Trends 2026, one of Bulgaria’s top conference formats that gather technology experts from various sectors. Joining experts from Google Cloud and Stihia.ai on the “When AI Faces AI” panel, he shared how the company applies this thinking through AI agents developed by its own security team. For a gaming software and technology supplier with more than 1,300 team members, international offices and over 200 clients worldwide, coordinating security across borders is a substantial undertaking. Popov outlined how Amusnet approaches that complexity while supporting innovation and strengthening the confidence of its partners.
AI backed by specialist expertise
Amusnet’s information security team develops and actively uses its very own AI agents to investigate cyber-attacks. The agents help correlate signals across identities, devices and cloud environments, bringing events and evidence into a clearer picture. This gives security experts more time to verify findings, assess risk and focus on decisions that matter to the business.
“AI is changing the rules of cybersecurity, but the strategic advantage is in our hands. Technology gives us powerful tools to build effective defences. At Amusnet, we develop our own AI agents, combining the speed of technology with the experience and judgement of our experts,” said Popov.
The team also uses these capabilities to assess whether newly reported cyber incidents elsewhere could reveal similar risks within Amusnet. Popov emphasised that when we talk about cyber-attacks, defenders have an important advantage: detailed knowledge of their organisation and its critical assets. AI helps specialists apply that knowledge more quickly.
Security that enables responsible innovation
For Popov, greater AI autonomy requires clearly defined limits. Organisations need to establish what agents can access, which actions require approval and how activity can be traced or stopped. Human oversight is effective when experts have enough context and a real opportunity to challenge a proposed action.
“Defining the task and the goal is not enough. You also need to limit the means an agent can use to achieve them,” he explained. Building those safeguards into projects from the outset can also help teams adopt new technology faster. Clear rules on data use, access and recovery give developers an established framework, reducing the need to resolve the same fundamental questions with each new initiative.
Trust that supports international growth
For Amusnet, which works with clients across more than 35 regulated markets, the business value of this approach extends to its international partnerships. Clear, verifiable answers about data access, system behavior and incident response give clients a stronger basis for confidence in the technology and the company behind it.
“For a company with a strong international presence like Amusnet, this approach is also important for the trust of partners and clients,” Popov said. Amusnet continues to invest efforts and expertise in turning its own technological know-how into business value by combining AI development with accountable security decisions. Building security into innovation supports the confidence to grow and the trust that sustains international partnerships.
