WIRED tests an AI agent worth the risk

A WIRED reporter tested an AI agent that saved $550, booked restaurant reservations and flagged a phishing scam, but also wasted $64 and may pose a security risk.

• Chatbots Are Leading Hikers Astray, With Dangerous Consequences2026-11-01
• Apple CEO: No AI for AI's sake2026-10-31
• WIRED Tests the AI Death Clock App2026-10-29
• AI agents manage fantasy soccer team, still trail humans2026-10-28
• Macmillan staff say publisher threw party for discontinued AI assistant2026-10-28