Hungarian AI landscape 2025: The strategic assessment of corporate AI adoption in Hungary

Deloitte's 2025 survey consolidates the experiences of 109 Hungarian organisations and documents precisely the inflexion point at which domestic AI transformation currently stands: the technology has become widespread, but strategic institutionalisation remains elusive for most organisations.

85% of participants are actively using some form of AI solution, and 83% plan to increase their spending in this area over the coming year. The intent to invest is clear — the question is whether this energy can be converted into structured business value or will dissipate across pilot programmes.

1. Strategy and organisational governance

The greatest weakness in domestic corporate AI adoption is not technological but organisational. Only 21% of organisations have a dedicated AI strategy, and 37% have no executive whose sole responsibility covers AI initiatives. This means the majority of projects are launched and run in an ad hoc manner, with unclear accountability — even as investment intent remains strong.

The survey data clearly show this is not a sustainable position. Successful AI implementations are consistently underpinned by the same three factors:

• a well-chosen use case that solves a measurable business problem
• active commitment and visible support from senior leadership
• deliberate development of internal competencies

The weight of senior leadership support has jumped ten percentage points in the success factor rankings since 2023 — the fastest shift in any measured dimension. The evidence is clear: where leadership does not take explicit ownership, AI investment becomes fragmented, and returns remain immeasurable.

The absence of an AI strategy is not a technology problem. Where senior leadership fails to set direction and assign accountability, investment fragments — and returns cannot be measured.

The most important step is developing a dedicated AI strategy with defined priorities, assigned owners, and measurable targets. Where immediate capacity for this is lacking, establishing an AI working group — bridging business units and IT — is a solid starting point. Successfully scaling one or two demonstrable use cases builds the executive confidence required for larger programmes.

2. Use cases and technology priorities

Domestic organisations are currently using AI primarily for what is easiest: automating internal processes (69%), improving resource allocation (62%), and replacing routine tasks. This is an understandable first step — it delivers quickly measurable results and requires little organisational transformation. The strategic risk, however, lies precisely here: organisations that think only in terms of optimising existing processes will miss the value creation opportunities that innovation-oriented applications offer.

On the technology side, generative AI dominates with 85% penetration, alongside established presences in NLP (58%) and machine learning on structured data (43%). RPA remains a proven baseline solution in operational areas. The greatest expectation over the next one to three years is agentic AI — 51% of organisations see strategic potential here. This is the technology where AI autonomously executes complex, multi-step processes without human intervention.

The combination of generative AI and RPA currently represents one of the fastest paths to ROI. The emergence of agentic AI signals, however, that process redesign will soon become a strategic imperative.

When selecting use cases, it is worth moving beyond automating individual process steps and targeting entire process segments. The best initial projects are those of moderate complexity that have a measurable impact on business outcomes — such as decision support or automated reporting. For agentic AI, small-scale pilots are advisable at this stage: organisations need to gain experience with autonomous systems before the market expects their routine use.

3. Governance and legal compliance

Governance is the area where domestic organisations are least prepared — and where risk is most tangible. 75% of respondents say their organisation either lacks or only partially has a comprehensive AI governance framework. Only 17% have an AI steering committee or ethics board, and a third of AI-related risks receive no treatment at all.

The EU AI Act leaves little room for deliberation. The regulation introduces concrete deadlines and sanctions:

• Screening out prohibited AI applications and developing employee AI literacy are mandatory requirements
• Non-compliance with provisions governing high-risk AI systems is subject to sanctions

Yet 38% of the organisations surveyed cannot even assess whether their solutions fall within the scope of the regulation

This is not an administrative problem — it is a business risk. Retrofitting compliance is more expensive and operationally more difficult than embedding it at the design stage. Furthermore, the absence of governance frameworks is a matter of trust: without transparency and accountability, the adoption of AI systems — among both internal users and customers — remains limited.

The first and most urgent step is identifying and classifying all systems within the scope of the EU AI Act. Where high-risk applications are involved, initiating the compliance process cannot be deferred. A foundational AI governance document — setting out ethical principles, risk assessment methodology, and monitoring mechanisms — is not a bureaucratic burden but a prerequisite for responsible scaling.

4. Organisational competencies and training

The fastest-growing barrier to AI adoption is the lack of employee competencies: this factor has risen from 24% to 42% since 2023. The explanation is straightforward — the more AI solutions go live, the more visible internal capability gaps become. Even the best tool is worthless if an organisation cannot interpret, operate, and critically evaluate its outputs.

The majority of domestic organisations (68%) choose internal upskilling of existing roles over external recruitment — the most cost-effective and sustainable path. However, the depth of training programmes varies considerably:

• 33% have a comprehensive internal training programme
• 30% offer e-learning modules
• 21% provide no AI awareness training at all

The last point is particularly problematic: the EU AI Act treats ensuring employee AI literacy as a mandatory requirement — deferring training therefore also constitutes a compliance risk.

Competency development is not an HR task — it is a strategic investment. Organisations that upskill their teams now will be able to scale their next AI programme 12 to 18 months sooner.

Training programmes should be designed with role specificity in mind: senior leaders need to understand strategic decision-making and EU AI Act obligations, business units need to grasp use cases and outcome evaluation, and technical teams need implementation and risk management expertise. The EU AI Act sets a minimum standard — a more comprehensive training architecture represents a genuine competitive advantage.

5. Competitiveness and next steps

61% of domestic organisations rate themselves as competitive or leading in AI adoption — a confident self-assessment. The data, however, paints a more nuanced picture. Nearly a fifth of organisations do not measure the impact of AI initiatives through KPIs, 79% of AI projects merely meet expectations, and only 10% exceed them. By internal benchmarks, organisations consider themselves successful — but this partly reflects a low level of ambition.

Most domestic organisations have already moved beyond the first phase of AI adoption: the technology has been adopted, and experiments have been run. The next phase is institutionalisation — where AI becomes embedded in strategic decision-making, process redesign, and organisational culture. The advantage that early movers gain at this level is durable: it does not stem from access to technology, but from data, experience, and process maturity.

Over the next three years, competitive positioning will not be determined by whether an organisation uses AI. What matters is whether it can embed AI in a structured, institutionalised way — into its culture, its decision-making, and its innovation processes.

Three areas warrant immediate action. 

First, measurement culture: every active AI programme should have KPIs assigned and regular business reviews introduced. 

Second, focus: rather than scattering resources across pilot programmes, proven and valuable use cases should be scaled. 

Third, preparation for agentic AI: organisations that begin redesigning critical processes around autonomous agents now stand to gain a 12 to 24-month market advantage.

How can Fluenta One enhance your AI strategy?

Fluenta One's AI-native process automation platform offers an answer to the strategic challenge posed by the survey findings: how can an organisation move from experimentation to an institutionalised, measurable AI transformation? The platform integrates autonomous agents directly into business processes — from data capture through approval to reconciliation — and adapts to each client's unique workflows.

Get in touch to find out how you can take the next step in your digital transformation journey.

Source: Deloitte Magyar AI Körkép 2025 — Domestic Survey on the AI Ecosystem, Challenges, and Success Factors

‍