A legal action filed by New York accuses TikTok of conducting deceptive experiments on minors by replacing functional safety tools with ineffective placebo versions. According to the state's lawsuit, thousands of teenagers and children were led to believe they had refreshed their feed algorithms to block harmful material, when in reality the platform left their content streams completely unaltered.
The Illusion of the Algo Refresh Tool
The core of the allegation centers on an internal trial that enrolled thousands of accounts, including minors. These participants attempted to activate protective utilities such as the "Algo Refresh" feature. In standard operation, this tool provides an essential safeguard by wiping existing recommendation history and halting the delivery of unwanted, distressing, or potentially toxic videos.
While young participants believed the toggle had reset their algorithm, the platform quietly suppressed the reset mechanism for those in the test group. Users outside the experimental cohort received the legitimate, functioning version of the tool and saw their feeds instantly change, whereas the experimental group continued receiving the same recommended material without any actual protection.
Mounting Legal Battles Across Multiple States
The action initiated by New York forms part of a broader nationwide legal offensive involving more than two dozen state lawsuits. These actions collectively accuse the tech giant of intentionally engineering an addictive interface for young demographics while issuing misleading assurances regarding user safety and well-being.
Defending its development practices, the company stated,
"We continuously make improvements to our platform which means that, like most companies, we regularly test new products and features to validate the experience and understand how they work in the real world."
Wider Pattern of Algorithmic Experimentation
The complaint highlights that this incident was not an isolated event. Earlier disclosures revealed another testing program in which safety features were withheld from 15 million users across the United States. That earlier intervention notably included a teenager who later tragically died by suicide, intensifying scrutiny around how social algorithms experiment with vulnerable audiences.
