This weekly update focuses on how businesses and organizations are building trust between humans and AI agents. As AI becomes more common on the internet, experts say we need new systems to make sure AI agents act responsibly and safely.

More than half of all internet traffic is now created by bots and AI agents, not just people. This big change means we need to rethink how we keep things secure online. Companies are realizing that knowing who is doing something and what they're allowed to do is super important when AI agents make decisions.

One of the biggest challenges is something called identity and access management. This basically means keeping track of which AI agents can access what information and what actions they're allowed to perform. Security leaders say the hard part isn't making AI agents work—it's understanding what they're doing and making sure they follow the rules.

In the world of customer service, AI agents are already helping companies by answering questions and solving problems faster. Companies are discovering that AI agents can handle up to 90% of regular customer questions, which saves money and helps people get answers quicker. However, these AI agents still need humans to step in when things get complicated.

Research from universities including MIT shows that the real challenge with AI agents isn't writing smart computer code—it's making sure the entire organization is ready for them. This includes making sure data is organized, costs are tracked properly, and everyone knows who is responsible if something goes wrong. Experts say that for every hour spent perfecting the AI, companies should expect to spend about four more hours getting it ready for real-world use.