Taiwan Ministry Labels Amap/Gaode Maps A Security Risk

Reports suggest Anthropic's Claude chatbot may have had unauthorized access capabilities from its initial release. This follows recent discussions regarding the necessity of securing AI bots before widespread deployment. The company noted that while the ability to detect vulnerabilities is a feature, it also presents a risk for security professionals. This marks a rare instance of the company publicly warning about the capabilities of its own products. Security experts have identified real risks regarding the use of artificial intelligence capable of controlling devices. According to documents from Anthropic developer of the Claude application, the success rate of hacking attacks via hidden text is 3.20 percent. The report suggests these vulnerabilities could allow for unauthorized device control. An International Labour Organization (ILO) report reveals that millions of people spend hours daily in degraded work environments. These psychosocial risks pose health dangers to workers. The report states these conditions cause hundreds of thousands of deaths every year. Marie-Uss-Lafond reported on the findings. Taiwan's Ministry of Digital Affairs has identified the Chinese navigation app Gaode Maps as a threat to national information and communications security.