User acceptance of AI is very important for using AI technology successfully. Recent studies show that acceptance depends on factors like how useful the user thinks AI is, what they expect from its performance, their attitude, trust, and how much effort it takes to use the technology. These factors mostly predict if a user will use or buy AI products. In healthcare administration, these things affect whether AI systems like Simbo AI\u2019s front-office phone automation become fully used or mostly ignored.<\/p>\n
The review by Sage Kelly, Sherrie-Anne Kaye, and Oscar Oviedo-Trespalacios found that while AI can be useful and sometimes easy to use, culture and the need for human contact can limit AI use. This is important in U.S. healthcare where patient interaction and personal service are very important. Therefore, AI systems must balance automation with chances for human contact.<\/p>\n
The extended Technology Acceptance Model (TAM) is often used to study what affects AI adoption. TAM looks at how useful and easy to use people think AI is. But trust and attitude towards AI are also key. People who trust AI and have a positive attitude toward it are more likely to use AI, even if it takes some effort to learn.<\/p>\n
Even though there is more interest in AI acceptance, many studies have weaknesses that affect their results. The review pointed out that most studies use self-reported data from surveys or interviews. These methods help at first, but they cannot capture how users interact with AI in real-time or real settings.<\/p>\n
Self-reported data can be biased. For example, people may give answers they think others want to hear (social desirability bias). They may also forget details or misunderstand what AI means. The review found that 31 out of 60 studies did not even clearly explain what AI means, which causes confusion and makes acceptance hard to measure correctly.<\/p>\n
To fix this, the authors suggest using naturalistic methods. These methods watch users as they use AI in their normal workplaces. This can include watching how they work (ethnographic observations), studying data logs from the system, or running tests that record real behavior instead of just intentions. For healthcare administrators using AI phone systems, naturalistic research can show real problems like issues in workflow, frustrations from users, or challenges in how humans and machines work together.<\/p>\n
<\/p>\n