Content Moderation

Core Summary BBC reports that researchers have discovered specific prompts can bypass ChatGPT’s safety filters to generate violent and sexualized images. This finding has reignited public discussion about the safety boundaries of AI-generated content and highlights the ongoing challenges large language models face in content moderation. Event Details According to BBC Technology, multiple independent research teams testing OpenAI’s latest image generation capabilities found that despite multiple built-in safety protections, carefully crafted indirect prompts can still induce the model to output content that violates usage policies. This content includes images depicting violent scenes and sexual suggestion. ...

Meta Repeatedly Refuses EU Requests Over User Ban Transparency EU regulators have disclosed that Meta has repeatedly failed to provide detailed information about user ban decisions on Facebook and Instagram, marking the third time this year the company has been cited for non-cooperation under the Digital Services Act (DSA). The EU’s Digital Services Coordinator Network (DSC Network) stated that Meta did not comply with DSA Article 24 requirements to share data on banned accounts within mandated timeframes. This data includes the number of banned accounts, ban reason categories, and appeal outcomes. ...

UK regulator Ofcom has reported that TikTok and YouTube are not safe enough for child users, calling for stronger content moderation and protective measures. The report notes that while both platforms have taken some protective steps, significant gaps remain in preventing children from accessing harmful content. Ofcom is requiring platforms to invest more resources in minor protection and consider implementing stricter age verification mechanisms. Source: BBC