Another patent from Sony aimed at enhancing the security of the PlayStation Network has surfaced.
While game developers usually file patent applications for technologies that directly benefit players, Sony’s latest filing appears to go beyond this scope, as its aim is to provide protection rather than a new gaming experience. Sony Interactive Entertainment has filed patent application No. 18/314775, which is intended to protect players from cheaters, potential child predators and other malicious individuals by proactively identifying suspicious users. The application is based on AI and essentially functions as a decoy account system. Published on August 13, the application aims to detect potential child predators, scammers, spammers and fraudsters.
“A pseudo account is set up to mimic the demographic and messaging behavior of a potential victim while seeming completely legitimate to an anti-social (AS) user. This can entail mimicking the messaging behavior of minors and presenting that behavior where AS users look for it. The AS account establishes communication with the pseudo account. The AS account may then be flagged as suspicious and the system can wait for the AS user to initiate contact with a legitimate account and report it before any undesirable communication takes place. The pseudo account can be controlled by a large language model (LLM), such as a generative pre-trained transformer (GPTT).” the summary says.
Put simply, imagine the patent as an undercover police operation within the context of an online game, controlled by AI. The patent’s goal isn’t necessarily to punish someone simply for chatting with an AI-supported account; rather, the system interprets this interaction as a signal that an account may be dangerous and then warns or protects real players when that account tries to contact them. Traditionally, game companies would have had to wait for someone to report a suspicious player. For example, a malicious player might contact a child and say inappropriate things to them. The child would then report the player and the platform would investigate the matter.
The purpose of the patent is to use AI to detect suspicious behavior when a malicious user approaches and flag the account, so that a warning appears when the malicious user approaches a real child. The system is designed to analyze how a given user interacts with the AI. If the user begins to exhibit suspicious behavior, the system flags this internally. Such behaviors include trying to move the conversation to a private platform, asking for personal information, trying to contact someone who appears to be a minor, sending inappropriate messages, attempting to deceive the account, sending spam and behavior associated with fraud. The account is not immediately suspended, but continues to be monitored. While the document specifically mentions minors and potential child victims, the system can also be applied more broadly to spammers, scammers, game cheaters and other antisocial users.
Bear in mind that, as with all patents, not every idea filed actually comes to fruition. Companies often patent ideas that they never bring to market. This simply indicates that the company is exploring the idea.
Source: MP1st
