Meta's AI bots threaten privacy and manipulate public opinion

By Jurica Dujmovic

Social media is increasingly dominated by artificial voices, with Meta Platforms (META) at the forefront of this transformation. The company has replaced its third-party fact-checking program with a community-based approach while simultaneously deploying thousands of AI personas across Facebook, Instagram, and Threads. This dual strategy raises alarms about the potential for AI to shape public opinion subtly and effectively.

The Perfect Setup for Manipulation

Meta's approach is straightforward: replace fact-checkers with community moderation, similar to X's (formerly Twitter) Community Notes system. However, Meta's Oversight Board has warned that this rushed rollout lacks proper human-rights due diligence. Meanwhile, AI bots are already infiltrating Facebook groups, blending into human conversations—sometimes with disturbing results. For instance, celebrity-voiced Meta chatbots were found engaging in inappropriate role-play with accounts posing as minors.

Key concerns:

• Unwanted AI Integration: Users are resisting Meta's AI push, as seen in the viral "Goodbye Meta AI" campaign, which garnered over 600,000 shares.
• Narrative Control: Meta gains unprecedented power to shape public opinion by deploying AI agents that mimic human interaction.

Behind the Curtain: Meta's Troubling Track Record

Meta's AI personas are not simple spam bots; they are powered by sophisticated language models capable of nuanced conversations. This capability is particularly worrying given Meta's history of exploiting user data, from the Cambridge Analytica scandal to psychological experiments on users. Despite promises to label AI-generated content, Meta's past misconduct casts doubt on its assurances.

Broader trends:

• Users across platforms like Reddit, YouTube, and 9GAG report increasing encounters with bots, blurring the line between human and artificial interaction.
• The "Dead Internet Theory," once dismissed as conspiracy, now seems plausible as tech companies openly flood the internet with AI-generated content.

X's Contrasting Approach

Unlike Meta, X (under Elon Musk) has taken a more aggressive stance against bots, focusing on spam and harmful content. X's tech-savvy user base and Community Notes system, which requires diverse viewpoint agreement, offer a more resilient environment against AI manipulation.

The Problem-Reaction-Solution Cycle

Meta's AI proliferation creates an artificial crisis, paving the way for invasive "solutions" like digital authentication systems that demand personal data. This shift threatens anonymous speech and privacy, all under the guise of combating a problem Meta itself engineered.

Regulatory gaps:

• Global authorities are focused on Meta's past infractions, missing the more pressing threat of AI-driven opinion manipulation.

A Call to Action

To counter this dystopian trajectory, the article proposes:

1. Shift the verification burden to companies deploying AI, not users.
2. Protect anonymous speech as a fundamental right.
3. Enforce transparency about AI's role in content creation and moderation.

Read more:

• What is AI really giving back to tech investors? Here's the hard truth.
• Dark-web AI models could make criminal hackers even more powerful

—Jurica Dujmovic
(END) Dow Jones Newswires
05-23-25 1648ET