Meta removes controversial Instagram AI feature after user backlash

It lasted only a few days.

Meta has rolled back a controversial Instagram AI feature after users questioned why the company was allowing AI to reference public Instagram content for image generation. The reversal came less than a week after Meta introduced the feature as part of the rollout of its new Muse Image model.

The feature aimed to make AI image generation feel more personal.

Users could ask Meta AI to create images inspired by public Instagram accounts, allowing the chatbot to draw on publicly visible photos as a creative reference. Meta said the tool relied on existing privacy controls and that users could manage whether their public content was used.

For many users, however, those assurances came too late.

Screenshots of the feature spread quickly across Instagram, Threads and X, with users questioning whether they had knowingly opted in and whether their photos could become training material or visual references for strangers. Privacy advocates argued that the issue was less about whether controls existed than whether most people understood them.

As criticism grew, Meta quietly removed the feature.

The company confirmed to TechCrunch that it had paused the rollout after listening to user feedback, saying it wanted more time to refine the experience before deciding whether to reintroduce it. It did not say when, or if, the feature would return.

The retreat comes at a sensitive moment for Meta.

Over the past week, the company has launched Muse Image, its new AI image-generation model, and Muse Spark 1.1, a coding-focused AI system aimed at developers. Together, the releases signal a broader strategy to weave specialised AI tools into Meta's products rather than rely on a single chatbot.

That strategy has also brought renewed scrutiny.

Across the technology industry, companies are increasingly racing to build AI features directly into consumer apps. At the same time, regulators in Europe and privacy campaigners have intensified scrutiny of how personal data, public posts, and user-generated content are used to power those systems.

Meta is not the first company to adjust course after an AI rollout.

Google has previously paused AI-generated search features after factual errors, while OpenAI has withdrawn or modified several ChatGPT capabilities in response to feedback from users and safety researchers. This pattern reflects a broader shift across the industry, where companies are increasingly releasing AI products quickly, monitoring public reaction and refining them after launch rather than waiting for lengthy development cycles.