This weekly update shows that trust between humans and AI agents is becoming the biggest challenge in business today. A major study found that trust in AI agents dropped from 43% to just 27% in one year.

Companies are worried about letting AI agents make decisions on their own. Only 2% of businesses have successfully used AI agents across their whole company. The main problems are concerns about privacy, unfair treatment, and not understanding how the AI makes choices.

However, some companies are finding ways to build trust. Kyndryl created a new system that keeps humans in control while letting AI agents do their work. They make sure every action by an AI agent can be tracked and explained.

Anthropic released a new AI model called Claude Opus 4.1 that is better at coding and complex tasks. They also shared rules for making AI agents that people can trust. Their main idea is that humans should always stay in control of important decisions.

Real companies are starting to test these systems carefully. A government and a bank are trying Kyndryl's system to see if it works for them. In healthcare, AI agents are helping with medical research but doctors still make the final choices.

The key lesson from this week is that trust must be built into AI systems from the start. Companies that do this right are seeing better results than those that add safety features later.