Only one-third (34%) of companies in Romania have an AI strategy approved by management or prepared for implementation over the next three to five years, compared with half (51%) in Central and Eastern Europe (CEE), according to the PwC survey ‘AI in CEE: From strategy to scale’. Nevertheless, interest is strong: more than half are working on a strategy expected to be finalised within the next 12 months, the highest share in the region (compared with over one-third in CEE). Only 10% have no strategy and no plans to develop one in the coming year.
“In Romania, the issue is not a lack of interest. The gap with the region emerges when plans need to be converted into results, amid smaller budgets, projects that remain in the pilot phase and very few internal rules governing the use of AI. However, the delayed start may also be seen as an advantage, as Romanian companies can build from the outset the foundations AI needs to operate at scale: high-quality data, skills and governance that accelerates implementation rather than slowing it down,” said Gabriel Voicila, Partner and Business Advisory Services Leader at PwC Romania.
The difference is also reflected in budgets allocated over the past 24 months to AI and advanced data analytics projects. In Romania, 67% of companies invested less than EUR 25,000, compared with 37% across the region. In the EUR 25,000 – 99,999 range, the shares are similar: 22% in Romania and 24% in CEE. The gap emerges in medium-sized budgets, between EUR 100,000 and 749,999, which account for 33% of companies in the region but only 6% in Romania. At the upper end, from EUR 750,000 upwards, the share is the same: 6%.
The greatest challenge remains moving from pilot projects to effective use
Nearly three-quarters of companies in Romania successfully implemented no more than one-quarter of the AI projects initiated over the past 24 months, the highest share in the region, where the average is close to half. Only 9% implemented more than half of the projects launched during this period, compared with one-quarter of companies in CEE. The results reflect this: nearly one-quarter say they achieved all or most of the expected benefits, compared with 40% across the region.
Where AI projects have been implemented, the benefits are most visible in IT, for nearly half of companies, and in finance and accounting, for more than one-third, the highest level in the region. The main reported effect is lower operating costs (27%). However, the most common answer among Romanian companies when asked about the impact of AI is “no impact observed”: 29%, almost double the regional share (16%). Growth effects are even less common: only 2% say AI has generated new revenue streams and 4% report a competitive advantage, compared with 13% and 14%, respectively, in CEE.
“It is no coincidence that AI creates value first in IT and operations. These functions have structured data and processes designed to be measured. The first results are better data quality and lower operating costs. However, these efficiency gains have their limits. The greater opportunity lies in new revenue, new products, and stronger competitive positioning. Most organisations in Central and Eastern Europe have not yet crossed this threshold,” said Anda Rojanschi, Chief Markets Officer, PwC Central and Eastern Europe.
Companies in Romania report fewer obstacles than those in the rest of the region. One-quarter say they face no organisational barriers, while more than one-third report no technological barriers, compared with 7% and 12%, respectively, in CEE. This indicates an earlier stage of adoption rather than an absence of challenges. The most frequently cited obstacles also remain below regional levels: high implementation and maintenance costs (38%, compared with 44%), lack of skills (30%, compared with 37%), and insufficient technological maturity (28%, compared with 33%).
The largest gap is in governance
Three-quarters of companies in Romania have no form of formal AI governance, compared with just over one-third across the region, and none report having a standardized model or one applied across the entire organisation. At regional level, 18% of companies already have such models.
To narrow the gap between strategy and execution, the survey identifies five approaches shared by organisations that successfully scale AI: treating AI as a transformation agenda, rather than merely as a cost-cutting tool; building the foundations—data, infrastructure and governance—before scaling; focusing on a limited number of priority projects; developing governance alongside adoption rather than afterwards; and extending validated use cases across teams and functions.
About the report
The PwC report “AI in CEE: From strategy to scale” is based on a quantitative survey conducted in the Czech Republic, Poland, Romania and Slovakia, covering 366 organisations with more than 250 employees.
