Research: Screening technology in business is limited by data quality, infrastructure, and unwillingness to change processes. The study covered more than 100 executives responsible for the AI ​​agenda in companies. In addition, in-depth interviews were conducted with IT leaders of large companies from industry, development, retail, pharmaceuticals and the agricultural sector, as well as surveys of technology providers. According to the survey, 43% of companies are experimenting with AI and pilot project management, another 31% are already using solutions in regular operation. The technology providers assess the situation more accurately: according to their estimates, only 10-15% of the large holdings have a full-scale productive implementation. These indicators belong to different groups: 31% - the share of respondents who use solutions in regular operation. 10-15% - estimate of the supplier, which applies only to large holdings and full-scale production performance. 65% of the respondents encountered pilots who did not reach an industrial operation: mostly due to the lack of a clear management decision and the reluctance of business process owners to change the usual regulations. At the same time, as the researchers point out, stopping the pilot is not necessarily a failure: closing the time of the project allows you not to invest resources in a solution that does not give the desired effect. Fences vary depending on the size of the business. In small and medium companies in the first place are imperfect models - this is what 62% of the respondents said. The quality of corporate data is often more stressed by large businesses - 83% of large organizations named this problem. Behind this, as a rule, there are deeper limitations. Unpredictable infrastructure and information security requirements. Russian details also play a role: according to the study, development is concentrated around 3-4 large ecosystems and companies rely on deployment in a secure loop.

Research: Screening technology in business is limited by data quality, infrastructure, and unwillingness to change processes. The study covered more than 100 executives responsible for the AI ​​agenda in companies. In addition, in-depth interviews were conducted with IT leaders of large companies from industry, development, retail, pharmaceuticals and the agricultural sector, as well as surveys of technology providers. According to the survey, 43% of companies are experimenting with AI and pilot project management, another 31% are already using solutions in regular operation. The technology providers assess the situation more accurately: according to their estimates, only 10-15% of the large holdings have a full-scale productive implementation. These indicators belong to different groups: 31% - the share of respondents who use solutions in regular operation. 10-15% - estimate of the supplier, which applies only to large holdings and full-scale production performance. 65% of the respondents encountered pilots who did not reach an industrial operation: mostly due to the lack of a clear management decision and the reluctance of business process owners to change the usual regulations. At the same time, as the researchers point out, stopping the pilot is not necessarily a failure: closing the time of the project allows you not to invest resources in a solution that does not give the desired effect. Fences vary depending on the size of the business. In small and medium companies in the first place are imperfect models - this is what 62% of the respondents said. The quality of corporate data is often more stressed by large businesses - 83% of large organizations named this problem. Behind this, as a rule, there are deeper limitations. Unpredictable infrastructure and information security requirements. Russian details also play a role: according to the study, development is concentrated around 3-4 large ecosystems and companies rely on deployment in a secure loop.