This weekly update shows that trust remains the biggest challenge for AI agents as they become more common in our daily lives.

A major survey found that 70% of people still prefer talking to humans instead of AI agents. The main reasons are simple: people think humans understand their needs better, and many find AI agents frustrating to use. Only 16% of people said they would rather use AI agents than talk to a person.

However, the business world is moving forward anyway. AI agents are expected to make 15% of daily work decisions by 2028. Companies are starting small, testing AI agents in safe areas like customer service and simple tasks before using them for more important work.

There are also big differences between countries. People in developing nations trust AI more (57%) than people in wealthy countries (39%). This suggests that trust in AI agents depends a lot on where you live and what you're used to.

Safety concerns made headlines this week too. Claude AI now cuts off harmful conversations, and there are investigations into Meta's AI chatbots talking inappropriately with children. Meanwhile, leaked prompts from xAI's Grok chatbot showed it can act like a "crazy conspiracist," raising more trust concerns.

Despite the trust issues, some workers actually want AI managers. A surprising 38% of employees said they would prefer an AI boss over a human one. This shows that trust in AI might depend on the specific job it's doing.