Two years ago, a health insurance claim at one of Greece's largest insurers took an average of 14 days to process. Medical opinions, lab results, invoices, and policy details had to be manually checked and cross-referenced before payment was approved.
Today, that same process takes about one minute, with compensation paid out the next day. The change is driven by AI Flows, an artificial intelligence tool developed by the Greek company Archeiothiki. The system handles roughly 1,500 claims daily, reading documents, extracting key information, and verifying it against policy terms and company rules.
More importantly, it enables something previously impractical: checking 100% of cases rather than a sample. This allows insurers to spot potential fraud that would otherwise slip through unnoticed.
From hours to minutes in banking
Similar technology is reshaping banking. When assessing companies for loans, banks must evaluate environmental, social, and governance (ESG) criteria through around 300 questions. Staff used to hunt through hundreds of pages of corporate reports to verify answers—a task that took about 16 hours per company.
Now, AI systems can scan reports of 400 or 500 pages, locate relevant passages, and process them in a fraction of the time. The final ESG assessment influences the interest rate a company receives, making accuracy and speed critical.
Archeiothiki currently applies AI Flows to about 100 workflows for ten clients, mainly banks, insurers, and healthcare firms. But Greece is not alone.
Which jobs are most exposed?
According to a May 2025 report by the International Labour Organization (ILO) and Poland's NASK research institute, administrative and clerical roles are the most exposed to generative AI. Data entry clerks, accounting staff, and secretaries top the list. In high-income economies, roughly one in three jobs shows some degree of exposure.
AI has already entered European workplaces at scale. A study by the European Commission's Joint Research Centre, published in October 2025, found that 30% of EU workers already use AI tools, with office work leading the way.
The shift has a clear gender dimension. Recent ILO research shows that occupations where women are the majority are almost twice as likely to be exposed to generative AI as male-dominated roles—29% compared with 16%. The reason is the higher concentration of women in administrative and support positions.
"I will not hide it. There are jobs that are lost too," Archeiothiki executives told Euronews. But they paint a more complex picture. Automation is not always about doing the same work with fewer people; it often allows companies to tackle tasks that were previously impossible at scale.
One client, for instance, had 50 people working on a single process and still faced a year-long backlog. In other cases, where only a sample could be checked before, now the entire dataset can be examined.
The ILO agrees that exposure does not mean elimination. Few occupations consist solely of tasks that can be fully automated today. The most likely outcome, for now, is transformation rather than replacement.
The next frontier: public sector archives
In Greece, the next major opportunity lies in the millions of documents already digitised by the state. Archeiothiki's CEO, Andreas Papadakis, says his company has handled about 15% of the relevant digitisation projects for the government.
The problem is that much of this digital archive remains "blind." Physical documents were scanned and given basic metadata for search, but their content was not converted into structured, analysable data.
The first stage of digitisation turned mountains of paper into digital files. The next stage is already underway: AI systems that not only read and record documents but extract information, cross-check it, and execute entire workflows.
As this technology spreads across Europe, the back office is being redefined. For workers in administrative roles, the storm is real, but so is the potential for new tasks and new ways of working. The key will be managing the transition fairly—especially for the women who make up the majority of these roles.
