engadget.com web signal

Meta Threads AI Account Resists User Blocks

meta ai assistants consumer-ai ai-assistants social-media

Key insights

  • Meta's @meta.ai Threads account cannot be blocked by users, only muted or hidden via a 'Not interested' option.
  • The backlash generated over one million posts trending on Threads within hours of the limited five-country rollout.
  • Even filing a spam report against the Meta AI account was reported by users as failing to trigger a block.

Why this matters

Deploying an AI account immune to the platform's standard block feature establishes a two-tier system where corporate AI entities operate under different rules than human users or third-party accounts, a precedent other platforms will watch closely. For founders building AI-integrated products, Meta's swift public backlash is a live case study in how consent architecture -- not just capability -- determines user trust and adoption velocity. Regulators in the EU and UK, already scrutinizing Meta's data practices under GDPR and the Digital Services Act, now have a concrete example of a platform overriding standard user controls to serve an AI product, which could accelerate enforcement conversations.

Summary

Meta's decision to deploy an unblockable @meta.ai account on Threads has ignited one of the platform's largest user revolts since launch. The account, currently being tested in five countries, uses Meta's Llama Spark model to inject AI-generated replies into public conversations -- and users discovered almost immediately that the standard block function simply does not apply to it. Within hours of the rollout, "Users cannot block Meta AI" became a trending topic with over one million posts. Users attempting to file spam reports found even that workaround failed to produce a block. Meta's only offered tools are mute and a "Not interested" option -- controls that suppress visibility but do not prevent the AI from continuing to engage with public threads. Essentially: (Meta, Threads users) are in a direct standoff over who controls the composition of a conversation. - The @meta.ai account is live in five countries during the test phase, with no announced timeline for global rollout or feature parity. - Meta spokesperson Christine Pai confirmed block is unavailable, framing mute and hide as sufficient "management" tools. - Spam reporting, normally a blunt-force moderation tool available to all users, was also reported as non-functional against the account. Platforms that build AI presence directly into social graphs while stripping standard user controls are setting a precedent that will be cited in every future regulatory debate about consent and AI interaction.

Potential risks and opportunities

Risks

  • Threads user churn in the five test markets could accelerate if Meta does not add a block option before broader rollout, handing Bluesky and Mastodon a recruiting moment among privacy-conscious users.
  • EU regulators could cite the unblockable account as evidence of dark-pattern design under the Digital Services Act, opening Meta to enforcement action before global Threads AI deployment completes.
  • Advertisers running brand-safety-sensitive campaigns on Threads face reputational exposure if @meta.ai replies appear adjacent to their content in ways they cannot control or anticipate.

Opportunities

  • Bluesky and Mastodon can use this moment to publish explicit AI-interaction consent policies, differentiating on user control as a product feature to capture migrating Threads users.
  • Privacy-focused tooling companies (Ghostery, DuckDuckGo) and browser extension developers have an immediate opening to build Threads-specific filters that surface and suppress @meta.ai activity at the client layer.
  • Policy consultancies and DSA compliance firms advising platforms on AI-feature rollouts gain a high-profile cautionary case to sharpen their consent-architecture advisory offerings ahead of anticipated EU AI Act guidance.

What we don't know yet

  • Whether the five test countries were selected partly because they fall outside GDPR jurisdiction, and what the block-feature policy will be when Meta AI rolls out to EU users.
  • Whether Meta's internal user-retention or engagement metrics showed any measurable drop in Threads activity in the hours following the rollout.
  • What content-moderation or liability framework governs @meta.ai replies -- specifically whether Meta AI responses are treated as platform speech or user-generated content under Section 230.