Meta adopts Community Notes model, using AI to review

Meta is making a massive change to how it moderates content across its platforms. The company is scrapping its third-party fact-checking program in the United States and rolling out a Community Notes system.

Inspired by rival X (formerly Twitter), the program will allow users to flag misleading posts and add context directly. Meta wants to ditch what it calls “over-enforcement” and move back to its roots: free expression.

The change couldn’t come sooner for many users. Meta’s platforms, including Facebook, Instagram, and Threads, have often been criticized for being too strict, often censoring harmless content and stifling political debates.

Millions of posts were removed daily last year, yet the company admits 10-20% of these takedowns may have been mistakes. That’s a staggering number when you consider the scale of its operations. Meta’s CEO, Mark Zuckerberg, says the company wants to “undo the mission creep” and return to being a place where people can speak freely.

Community Notes hands the power to users

Meta will no longer write or control the notes or decide which ones are visible. Instead, users will draft and rate notes on flagged posts. To avoid bias, agreement from people with different perspectives will be required before a note is approved.

Meta promises to show how diverse viewpoints contribute to the notes and will expand its reporting on moderation errors to build trust. The rollout begins in the United States over the next few months. People can already sign up on Facebook, Instagram, and Threads to be among the first contributors.

Once live, Community Notes will replace Meta’s current content moderation tools, including the much-hated interstitial warnings that blocked posts until users clicked through. These will be replaced with smaller, less intrusive labels offering more context for those who want it.

In December 2024, Meta removed millions of pieces of content every day. While this represented less than 1% of all posts, the sheer volume of mistakes exposed flaws in the system. Users frequently complained about being thrown into “Facebook jail” or having their posts unfairly flagged.

AI to clean up moderation and boost efficiency

Artificial intelligence is at the heart of Meta’s overhaul. The company has been quietly integrating large language models (LLMs) into its moderation processes.

These AI systems provide a second opinion on flagged content before enforcement actions are taken. This has already reduced errors and sped up appeals for users who believe their posts were wrongfully removed.

AI is becoming a core part of Meta’s broader strategy. On January 1, the company announced plans to introduce AI-generated user-profiles and content across its platforms. According to Connor Hayes, Meta’s Vice President of Product for Generative AI, these profiles will function like real users, creating and sharing posts autonomously.

This decision has brought criticisms, but Hayes says it’s all about making platforms more dynamic and engaging. By late 2024, over 600 million users were actively using Meta’s AI-driven features each month. Chatbots in Messenger and advanced character creation tools have already been a hit.

These tools allow users to create AI personas that interact seamlessly with others. It’s a glimpse into Meta’s vision of blending human and AI interactions across its platforms.

Meta’s financial and strategic focus on AI

Generative AI tools have already transformed Meta’s advertising business. Since their launch, more than 15 million ads have been created using these tools, with companies leveraging AI to craft hyper-targeted campaigns.

Analysts believe these tools could open new revenue streams, especially in customer service. WhatsApp, for instance, is testing enterprise chatbots powered by AI, which could evolve into a multi-trillion-dollar market.

Investors are optimistic. Jefferies analysts recently raised Meta’s stock price target to $715 per share, citing confidence in its AI-driven growth strategy. Meta’s stock surged 70% in 2024, far outpacing the S&P 500’s 24% gain. This performance is tied directly to Meta’s ability to monetize its AI tools and keep users engaged.

Meanwhile, Zuckerberg is doubling down on AI across the board. At a recent conference, he revealed plans to embed generative AI into experimental products like Ray-Ban smart glasses and the Orion augmented reality headset.

Despite its successes, Meta’s AI strategy isn’t without risks. Regulatory hurdles in the European Union are slowing the deployment of its AI tools, with stricter rules around data usage and transparency. Critics also warn about the potential for misinformation and low-quality content flooding platforms. AI-generated posts, while innovative, could make it harder for users to separate fact from fiction.

Land a High-Paying Web3 Job in 90 Days: The Ultimate Roadmap