Organisations working with artificial intelligence and machine learning have an average of four projects underway, and plan to add 15 more within the next three years, according to a Gartner survey.\nThe small survey of 106 Gartner \u2018Research Circle\u2019 members, found about three in five of the respondents had AI deployed today.\nBy 2022, Gartner predicts the organisations to each have 35 AI-powered applications and projects in place.\n\u201cWe see a substantial acceleration inAI adoptionthis year,\u201d saidJim Hare, research vice president atGartner.\n\u201cThe rising number of AI projects means that organisations may need to reorganize internally to make sure that AI projects are properly staffed and funded. It is a best practice to establish an AI Centre of Excellence to distribute skills, obtain funding, set priorities and share best practices in the best possible way,\u201d he said.\nThe top two use cases for AI currently deployed was to improve decision making and recommendations, and process automation. About a third had a virtual assistant or chatbot, and 14 per cent had embedded AI in products.\nThe most common motivations for rolling out AI was to improve the customer experience and to automate repetitive or manual tasks. Cost reduction and revenue growth were also cited as motivators.\n\u201cIt is less about replacing human workers and more about augmenting and enablingthem to make better decisions faster,\u201d Hare said.\nAdopting AI comes with considerable challenges, respondent reported. The most common were a lack of skills (cited by 56 pre cent of those questioned), understanding AI use cases (42 per cent), and concerns with data scope or quality (34 per cent).\n\u201cFinding the right staff skills is a major concern whenever advanced technologies are involved. Skill gaps can be addressed using service providers, partnering with universities, and establishing training programs for existing employees,\u201d said Hare.\n\u201cHowever, establishing a solid data management foundation is not something that you can improvise. Reliable data quality is critical for delivering accurate insights, building trust and reducing bias. Data readiness must be a top concern for all AI projects,\u201d he added.