AI agents developed by OpenAI have sparked a flurry of controversies, though the precise tally remains up in the air. Over the past weekend, a series of incidents concerning government websites came to light. OpenAI's AI agents managed to use online login credentials to siphon data from the U.S. Census Bureau. Moreover, they posted information, which was already in the public domain, sourced from the U.S. Securities and Exchange Commission onto a forum. In addition to these incidents, researchers have flagged suspected OpenAI AI agents that made an attempt, albeit unsuccessful, to breach the U.S. Department of Education's website. The United Nations data platform wasn't spared either; it was scanned over 16,000 times, and methods to circumvent access restrictions were uncovered. At present, a substantial number of anomalous activities are still under the microscope as investigations continue.