Meta זורקת את כלי יצירת התמונות ימים אחרי ההשקה — הוליווד רותחת

Meta abruptly scrapped a new AI image tool just days after launch, following a wave of privacy criticism and Hollywood outrage over a feature that automatically accessed users' public Instagram images. The company conceded the feature 'missed the mark' — a rare public retreat that underscores how sensitive image-training and likeness rights have become.

The backlash was compounded by a separate Gizmodo report that Meta's own AI detector couldn't reliably flag images its systems had generated — an embarrassing gap given Meta's public commitments to labeling AI content. Together, the two missteps fueled a narrative of Meta stumbling on AI governance even as it pours capital into infrastructure, including a new Louisiana data-center project.

Mechanically, the pulled tool's problem was consent: auto-accessing public images for AI features without clear opt-in triggered both legal-risk concerns and creator backlash over unlicensed use of likenesses and styles. The AI-detector failure highlights the still-unsolved technical problem of provenance — generative models can't reliably detect their own outputs, undermining watermarking and labeling regimes industry-wide.

Competitively, this contrasts with the week's product-momentum stories from OpenAI, Anthropic and xAI, and reinforces community skepticism about Meta's AI execution — CEO Zuckerberg has internally acknowledged agent progress trailing timelines, and The Register reported Meta admitted its first 'superintelligence' attempt was too weak to survive three days. Watch whether Meta reintroduces the image tool with proper opt-in, and how the detector failure affects its content-labeling commitments to regulators.