Periodically IG conducts large-scale sweeps in which many accounts are suspended, then the outcomes are evaluated.
In other words, they test different algorithms; each round of suspensions is treated as an experiment that is subsequently analyzed.
When they are not actively banning accounts, they continue to collect data to assess how well a given approach performed.
Their algorithms are not infallible. False positives are common: ordinary users can be flagged because they connected via poor Wi‑Fi or unstable mobile networks that happened to be associated with spam activity, or for other incidental reasons.
As a result, the detection systems often produce erroneous suspensions. IG will sometimes disable an algorithm, refine it, and then re-enable it.
Accounts that are functioning today may therefore be subject to another temporary sweep in six months, a year and a half, two years, or five years; if the account later triggers the algorithm, it can be suspended.
A recurring and particularly serious problem has been the misclassification of benign family photographs. For example, users who posted images of their children years ago—vacation photos or casual family shots—have in some cases been accused of Child Sexual Exploitation (CSE) and had their accounts suspended, sometimes with no successful path to restoration.
There are numerous reports of parents finding their accounts permanently blocked and being left in shock. These outcomes reflect changing policies and evolving detection systems.
Because rules and algorithms change continuously, users and businesses must constantly adapt and navigate these restrictions to maintain account access and operation.