A 80% of managers of companies believes that automation can be applied to any business decision, as revealed by a survey conducted by Gartner. This is a survey conducted at the end of last year among 699 people from the United States, Germany and the United Kingdom, in organizations that have deployed Artificial Intelligence or intend to deploy it in the next 3 years.
This survey has been used for the elaboration of a study, which shows that companies are moving from a tactical approach regarding Artificial Intelligence and are beginning to apply it more strategically. Thus, a third of organizations are applying Artificial Intelligence in various business divisions, differentiating themselves competitively by supporting its use for decision-making in various business processes.
In the study it can be seen that, on average, the 54% of Artificial Intelligence projects go from pilot to production. This is a slight increase from the 2019 AI in Organizations Survey, which found that 53% of AI projects went into production, down 1%. 40% of the organizations surveyed also indicated that they have thousands of AI models deployed. This creates more governance complications for the organization, and challenges the data and analytics capacity of managers to demonstrate the ROI of each model.
Although talent shortages are often assumed to be a negative impact on AI initiatives, the survey found that it is not a notable barrier to AI adoption. 72% of managers said they have or can find the AI talent they need. Security and privacy concerns are also not a big deterrent to AI adoption, worrying just 3% of managers surveyed. Of course, 41% of organizations say they have been aware that they have experienced a security incident or an AI privacy breach.
As for which areas of the organization are most concerned about AI security, 50% of survey respondents point to concerns about competitors, partners, and various third parties. Another 49% were worried about malicious hackers. However, among organizations that have been confronted with an AI-related security or privacy incident, a 60% indicate that data was compromised by someone inside.
