Major software-focused operations in Hungary reported robust results for 2025: the companies examined recorded a combined revenue of 352 billion HUF, more than an 11 percent increase year-on-year. Their aggregate after-tax profit approached 19 billion HUF, an improvement of roughly 28 percent compared with 2024.
Which units were reviewed?
The analysis reviewed Hungary-based subsidiaries and development centers that primarily employ developers, programmers, engineers, IT specialists and consultants. The entities covered include:
- SAP Hungary Kft.
- EPAM Systems Kft.
- Ericsson Magyarország Kft.
- Lufthansa Systems Hungária Kft.
- Morgan Stanley Magyarország Elemző Kft.
- BlackRock Hungary Kft.
- Cloudera Hungary Kft.
These units work on enterprise resource planning and cloud services, bespoke software development and R&D for telecommunications and aviation, as well as financial‑technology and big‑data platforms.
Employment and profitability
Alongside revenue growth, employment also rose: the combined headcount of the examined operations reached almost 9,300 people, an increase of about 500 employees (roughly 6 percent) year‑on‑year. The 352 billion HUF top line was accompanied by nearly 19 billion HUF of after‑tax profit, with the aggregate result improving by about 28 percent relative to 2024.
Some owners opted to retain the year’s profits as reserves (for example at SAP and Cloudera), while other firms’ managements proposed dividend payments (including BlackRock, Morgan Stanley, Ericsson and EPAM). Taken together, the reviewed companies may propose about 28.6 billion HUF in dividends from the 2025 results.
No clear AI impact in the headline numbers yet, but risks are real
The newly published financials do not yet show that the AI wave has damaged the domestic software sector: revenues and employment grew in 2025. Nevertheless, the story is far from settled. The spread of large language models and other AI tools — particularly those targeted at programming — could fundamentally change the revenue structure of software companies over the medium term.
AI is already taking over parts of software work, especially repetitive tasks typically performed by junior developers. Since ChatGPT became public in November 2022, such models have made steady progress; a recent step change came with new programming‑oriented models released in February of this year. These systems can now produce complex code snippets and in some cases assemble complete programs from textual prompts.
Why business models may shift
Many enterprise software vendors (notably SaaS firms) generate revenue with user‑based licensing, the so‑called seat‑based model. If AI agents dramatically boost productivity, companies may require far fewer human "seats" — and therefore far fewer paid licenses. That dynamic can erode license revenues and force an overhaul of existing pricing models.
Markets have already started to react: investors trimmed valuations for some software and IT services companies that rely heavily on large pools of junior developers doing repetitive coding. While the current financial results do not imply an immediate collapse, they do suggest the possibility of a paradigm shift in the coming years.
What to watch next
Observers should monitor next year’s reports for trends in license revenue, cloud service fees, and headcount. For managers, key questions are how to embed AI into offerings without destroying revenue bases, and how to redesign pricing if seat‑based structures become ineffective.
In sum, the 2025 financials show solid performance for the reviewed software operations in Hungary, but the pace of AI development and its adoption could prompt significant industry change over the medium term.
