Meta Muse Exposed Users' Personal and Business Secrets Through Shared Conversations

·1 views

According to Ars Technica, a single query was enough to surface thousands of shared conversations from Meta's Muse, listed by their titles. The exposure reportedly revealed users' personal and business secrets once they had chosen to share their chats. Because the conversations could be found so easily, the flaw made private discussions accessible to anyone running the search. The source material available here is brief, so details such as the exact scope, timing, and Meta's response are not confirmed in this summary.

Read the full story at Ars Technica

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

TikTok Staff Flagged Child Safety Risks, Court Filings Say Executives Did Not Act

Court documents reportedly show that TikTok employees raised concerns about risks to children on the platform. The filings indicate that staff were frustrated with how company executives responded to those warnings. The records suggest the concerns were not adequately addressed by leadership. The report is based on material disclosed in the court proceedings.

0 comments Read more at TechCrunch

TechnologyThe Verge ·

Graded Pokémon and sports cards worth over $963,000 stolen from FedEx truck in Santa Ana

More than 200 packages of graded trading cards being returned to their owners were taken from a FedEx truck last week. The truck was parked outside the headquarters of Professional Sports Authenticator (PSA) in Santa Ana, California. PSA, a major third-party grading company for sports and hobby cards, reported the theft to the Santa Ana Police Department. The company says the majority of the missing packages are worth more than $963,000, though that figure may rise as PSA confirms exactly which cards were taken.

0 comments Read more at The Verge

TechnologyThe Verge ·

Spotify removes 50 million AI-generated spam tracks from recommendations

Spotify says it has taken more than 50 million songs it considers AI-generated slop or spam out of user recommendations over the past year. The company made the disclosure in an announcement on Tuesday, which also described changes to its recommendation system aimed at surfacing more music from human artists. Spotify states that its editors apply the same standard and do not program content identified as slop or spam. The company reports a marked decline in listener complaints about such content appearing in recommendations since the changes were made. Spotify has not required labels on all AI-made tracks but has introduced an "AI Persona" badge for some content.

0 comments Read more at The Verge

TechnologyTechCrunch ·

Figma's Dylan Field says MCP is not a perfect protocol

Figma co-founder Dylan Field has said that the Model Context Protocol (MCP) is not a perfect standard. The company says it intends to support as many standards as it can. Figma aims to do this without exposing all of its underlying data. The source article is brief, so details about timing, specific shortcomings, and the reasoning behind the position are limited. Readers should treat this as a short report on Field's comments rather than a full account of the debate.

0 comments Read more at TechCrunch