It’s a question that’s becoming increasingly common, isn't it? "Is that Instagram photo AI?" The line between reality and digital creation is blurring faster than ever, and platforms like Meta are grappling with how to keep us informed. We saw Meta make significant moves in 2024 to tackle the growing challenge of AI-generated content, aiming to bring clarity to our feeds.
Back in February 2024, Meta announced a shift in its approach to labeling AI-created content. This wasn't just a minor tweak; it was a response to a world where AI can be used for everything from harmless fun to deeply concerning scams. We’ve heard stories, like the one about Jake, who lost a staggering AUD 130,000 after being duped by a scam that leveraged AI to impersonate a celebrity. While this particular incident was a stark reminder of the risks, Meta's policy changes, though perhaps arriving too late for some, are designed to prevent future harm.
The core of Meta's evolving strategy involves adding text labels to photos and videos that are detected as containing AI elements. This applies whether the entire piece is AI-generated or just partially manipulated. It’s a continuous process, and Meta continued to refine its policies throughout 2024. A detailed blog post in April, titled "Our Approach to Labeling AI-Generated Content and Manipulated Media," underscored this ongoing commitment.
This push for transparency wasn't unique to Meta. In May, the social media giant began rolling out "Made with AI" labels across its platforms for AI-generated videos, images, and audio. This expanded on an earlier policy that only covered a limited scope of doctored videos. Monika Bickert, Vice President of Content Policy, highlighted that more prominent labels would also be applied to digitally altered media that poses a "particularly high risk of materially deceiving the public on a matter of importance," regardless of how it was created.
Looking ahead, the global regulatory landscape is also sharpening its focus. China, for instance, is set to implement its "Measures for the Identification of AI-Generated Content" starting September 1, 2025. This new regulation aims to foster healthy AI development, standardize AI content labeling, and protect the legitimate rights of citizens and organizations. It defines AI-generated content broadly, encompassing text, images, audio, video, and virtual scenes, and outlines both explicit (user-perceptible) and implicit (data-embedded) labeling methods. The measures detail specific requirements for how these labels should be applied across different media types, ensuring that even downloaded or copied content retains its explicit identifier.
Internationally, organizations like the UN have been instrumental in shaping discussions around AI governance, with resolutions calling for a coordinated approach to AI regulation. UNESCO and the World Intellectual Property Organization (WIPO) have also been active, with UNESCO's recommendations on AI ethics and guidelines for generative AI in education, and WIPO's report suggesting measures like "record-keeping" for AI tools. The European Union's AI Act, which officially took effect in August 2024, imposes strict transparency obligations, requiring AI systems to ensure natural persons know they are interacting with AI, and mandating machine-readable labeling for generated content. Violations can lead to substantial fines.
In the United States, a patchwork of federal and state legislation is emerging. Proposals like the "AI Disclosure Act of 2023" suggest explicit labeling, while NIST's report in April 2024 outlined methods for detecting and marking synthetic content. California's "Artificial Intelligence Transparency Act of 2024" even mandates implicit labeling within large AI model systems. The US government is also focusing on accountability, with executive orders and FTC guidelines emphasizing transparency, explainability, and fairness in AI systems.
It’s clear that the conversation around AI-generated content is far from over. As we move through 2025, expect these labeling efforts to become more sophisticated and widespread, driven by both platform initiatives and evolving global regulations. The goal, ultimately, is to empower us, the users, with the knowledge to discern what's real and what's been digitally crafted.
