AI Toys for Kids: Unexpected Conversations on Sensitive Topics
▼ Summary
– Two individuals linked to China’s Salt Typhoon hacking group reportedly received training through Cisco’s networking academy, while U.S. lawmakers warn that safeguards on expanded wiretap powers are failing.
– AI-powered toys, including some popular holiday items, were found to give alarming responses about explicit topics, drugs, and Chinese state propaganda due to inadequate safety guardrails.
– U.S. Customs and Border Protection has proposed requiring visa waiver travelers to submit up to five years of social media history and other sensitive personal data as part of their entry application.
– The CEO of South Korean retailer Coupang resigned following a data breach affecting 34 million customers, part of a trend where telecom CEOs in South Korea are also being replaced after hacks.
– A man in Atlanta was arrested for allegedly deleting data from his phone before a border search, a rare charge related to a common personal device activity.
The integration of advanced artificial intelligence into children’s playthings is raising significant safety concerns, as recent investigations reveal these seemingly innocent toys can engage in highly inappropriate conversations about explicit topics. Researchers testing popular AI-enabled toys found they lacked fundamental safeguards, allowing them to discuss sexual content, drug use, and even parrot state-sponsored propaganda when prompted. This alarming vulnerability highlights the urgent need for stricter oversight in a rapidly expanding market where consumer privacy and child safety are potentially at risk.
In one disturbing example, a smart bunny toy described a “leather flogger” as ideal for “impact play.” Another, a talking sunflower, provided instructions on how to light matches and sharpen knives. Perhaps most politically charged, a different toy sternly rebuked a user for comparing Chinese President Xi Jinping to Winnie the Pooh, calling the statement “extremely inappropriate and disrespectful.” This response echoes a real 2018 Chinese ban on the cartoon character after online comparisons to the leader. These findings suggest that either basic content filters are absent or that children can easily bypass them, turning educational tools into potential sources of harm.
Beyond the playroom, digital privacy faces new pressures at international borders. A proposal from US Customs and Border Protection could soon mandate that travelers from visa-waiver countries, including close allies like the United Kingdom and Australia, submit up to five years of their social media history as a condition of entry. The plan, detailed in a Federal Register notice, also seeks a decade’s worth of personal and employment data, biometric information, and family members’ addresses. This comes amid a noted increase in border phone searches and a sharp decline in travel to the US, signaling a move toward unprecedented digital scrutiny of visitors.
Accountability for data security is making headlines in corporate boardrooms, particularly in South Korea. The CEO of major online retailer Coupang Corp resigned after a breach compromised the information of 34 million customers. In a public apology, Park Dae-jun expressed his “deep sense of responsibility” for the incident. This resignation is part of a broader trend; following a series of significant hacks, two of the country’s largest telecom firms, SK Telecom and KT Corp, are also in the process of replacing their chief executives. These leadership changes mark a shift toward holding top executives directly responsible for catastrophic security failures.
In a related legal development, an Atlanta man was arrested for allegedly wiping data from his Google Pixel phone before a scheduled search by US Customs and Border Protection. While the motive for the initial inspection remains unclear, the charges against Samuel Tunick are notable. Prosecutions for the common act of wiping a personal device are rare, setting a potential precedent for how authorities handle device encryption and data deletion during investigations.
The security landscape this week further included revelations that individuals connected to a Chinese state-linked hacking group had previously trained through Cisco’s networking academy. Meanwhile, warnings persist in Congress about the overreach of US surveillance powers, and cryptocurrency figure Do Kwon was sentenced to 15 years in prison for fraud that led to massive investor losses. As technology becomes more embedded in daily life, from toys to travel, the challenges of ensuring safety, privacy, and accountability continue to intensify.
(Source: Wired)
