Reports of unsafe AI-toy responses renew calls for child-safety rules
-
Freshways milk-price cut deepens pressure on UK dairy farms
-
India climbs green-transition ranking after first emissions decline in 25 years
-
Labour and National remain virtually tied as New Zealand vote count advances
-
Trump names Stephen Miller as White House chief of staff after midterm losses
-
The Good Daughter adapts Karin Slaughter thriller with Rose Byrne and Meghann Fahy
-
Restored Albanian tower houses anchor new rural tourism trail
-
Heavy-lift drone evacuates wounded Ukrainian soldier from frontline kill zone
-
Sonia Gandhi reflects on family tragedy and reluctant political career
-
Morrison says West Brom recovered after pressure from successive derby losses
-
Pope Leo begins three-country South American tour in Uruguay
-
Britain considers 2027 state visit for Ukraine’s Zelensky
-
Khawaja says entire Australia camp knew of 2018 ball-tampering plan
-
Arizona Democrats Secure Historic Trifecta with New Legislative Leadership
-
Anna Miltina Confronted with Racism Allegations on ProSieben Show
-
Carney Welcomes U.S. Trade Reversal After Democratic Midterm Victory
-
Microsoft Xbox App Update for PC Promises 60% Faster Startup
-
Armenian FM Discusses EU Accountability for Palestinian Rights in Madrid
-
California Humanities Honors Valdez and Burroughs for Courageous Leadership
-
Trump and Vance Pursue Secret Diplomacy with Iran Amid Military Pressure
-
Friedmann and Kühner Analyze Bundesliga's Best Shot and Frankfurt's Goalkeeper Crisis
-
Classic Car Values Surge: De Tomaso Vallelunga Leads 190% Gain
-
Honoka‘a Dragons Face Hawaii Prep in BIIF Division II Championship
-
Microsoft Ends Oracle AI Computing Rental Deal Amid Cloud Strategy Shift
-
Pakistan Demands Grant-Based Climate Finance at COP31 to Avoid Debt Trap
Reports of unsafe AI-toy responses renew calls for child-safety rules
Investigations into conversational toys have found that some products can produce disturbing or unsafe responses to children, increasing pressure for stronger regulation.
A teddy bear marketed for children as young as three reportedly made an unprompted reference to killing after mishearing a question about babies. A separate 2025 test by the US PIRG Education Fund found three AI toys gave information about finding potentially dangerous household items including matches, knives and plastic bags.
One toy, later withdrawn from sale, reportedly explained how to light a match and engaged in extended discussion of sexually explicit subjects. The incidents demonstrate how speech recognition errors and weak content safeguards can expose children to material not intended by parents.
The market remains lightly regulated in some countries. The European Union has introduced new requirements, while the UK is consulting on its approach. Parents are being advised to review privacy practices, supervision tools, data retention and age-appropriate safeguards before purchasing connected toys.
S.Gantenbein--VB
