The same week that TikTok announced a major push to help users spot and understand AI-generated content, Meta found itself in the middle of a firestorm over automatically opting users into its new AI image tools. Taken together, the two stories offer a snapshot of where the industry stands on a question that keeps getting louder: what does responsible AI actually look like, and who gets to decide?
TikTok’s announcement, published July 10, centers on three initiatives. The platform is launching a new in-app AI literacy hub, built in partnership with the National Association for Media Literacy Education and deepfake expert Henry Ajder, that surfaces practical tips for identifying AI content when users search AI-related terms. TikTok also confirmed it is joining the C2PA Steering Committee, deepening its commitment to Content Credentials, the industry standard for attaching provenance data to media. The company says it has now labeled more than 3 billion videos as AI-generated through a combination of Content Credentials, creator labeling tools, and invisible watermarking.
The third piece targets what many users would call AI slop. TikTok says it will begin testing improved detection systems aimed at accounts dedicated to posting AI-generated spam, with a focus on sensitive topics like politics, financial advice, and health information. The platform is also testing a Manage Topics feature that lets users adjust how much AI-generated content shows up in their feed, effectively handing people a dial for their own experience.
Meta’s week went differently. The company launched Muse Image on July 7 across the Meta AI app, Instagram, WhatsApp, and the web. The tool lets users generate AI images that reference any public Instagram profile simply by tagging a username. The catch is that every adult with a public account was opted in automatically, and turning it off requires digging into the Sharing and Reuse section of Instagram’s settings. Accounts belonging to users under 18 are excluded, and Meta says the feature shipped with strong controls and safety guardrails from day one.
The backlash was immediate. Tech reviewer Marques Brownlee flagged the default opt-in, talent agency CAA raised concerns on behalf of clients, and SAG-AFTRA and influencer managers pushed back over likeness and deepfake risks. Public Citizen’s JB Branch called it an invasion of consumer privacy, arguing that burying an opt-out in settings is not meaningful consent. Digital rights groups are now calling for an explicit opt-in model, where users grant permission before their content can be referenced.
To be fair to both companies, neither approach is purely altruistic. TikTok benefits from being seen as the trustworthy platform while facing its own scrutiny over AI content flooding feeds, and Meta genuinely built controls, exclusions for minors, and reporting tools into Muse Image. The real difference comes down to defaults. TikTok’s announcements put choice and disclosure in front of users. Meta’s launch asked users to find the exit after they were already inside.
That distinction matters because defaults shape behavior at scale. Most people never touch their settings, which means the decision a platform makes on day one is the decision most of its users live with. As generative AI becomes standard equipment on every social platform, the consent question is not going away. This week just made it impossible to ignore.
Sources:
gfpods.com 
Leave a comment