For Hungarian small and medium-sized enterprises (SMEs), AI has shifted from experimental to essential for staying competitive. Yet many companies still rely on ad-hoc, unregulated use. The main danger is often not an AI error itself but the phenomenon known as Shadow AI: employees uploading sensitive data, proposals or decision-support information into various external systems without supervision or control.
In the latest episode of the Portfolio Business Podcast, Varga Botond, AI and Innovation Director at Serco Informatika Zrt., argues that the "let’s try ChatGPT" era will end by 2026, as both the market and regulators increasingly demand deliberate, structured AI operations. The episode is available on Spotify, Apple Podcasts and other major podcast platforms, and via the embedded player in the article.
What companies should implement: rules, approved tools, measurable goals
According to Varga Botond, the answer is not to ban AI but to put a system around its use. Concretely this means:
- clear internal policies and approval workflows,
- use of approved and monitored tools,
- defined core processes and responsibilities,
- and measurable targets and KPIs so organisations can quantify time and cost savings.
A step-by-step rollout that delivers early, small wins helps demonstrate business value and build trust.
Fast learning for busy leaders: microlearning
A practical challenge is that leaders and teams often lack time for long training sessions. The specialist recommends a microlearning approach: 5–10 minute focused mini-lessons with immediately applicable tasks, enabling even the busiest staff to progress without leaving their daily work.
What to watch for in 2026
The article stresses that by 2026 the question is no longer whether companies should use AI, but which ones have secure, controlled and profitable AI systems and which rely on a quick but risky workaround. It is essential to measure both business returns and data security risks for every deployed solution.
Related content and publication
Related pieces include “Már látják a magyar cégek, kitől és milyen munkát vesz el az AI” (2026. 04. 21.) and “Rengeteg céges AI-projekt bukik el, az adatbiztonság is aggasztó, de van megoldás a problémára” (2026. 05. 15.). The publication of this article and the podcast episode was supported by Serco Informatika Zrt. Cover image is for illustration; source: Getty Images.
Tags: SME, technology, digital, artificial intelligence, digital transformation, small and medium-sized enterprise, 2026, ai, innovation, Serco Informatika Zrt., Varga Botond
