Microsoft has filed patent application with the U.S. Patent and Trademark Office for machine-learning artificial intelligence that will be able to detect cheating in video games. The purpose of the artificial intelligence in question is to detect cheating outside the games themselves, since platforms like Xbox Live are unable to do so.
“A platform that hosts third-party games may not be able to detect cheating that occurs in third-party games, even where achievements in third-party games are managed at the platform level,” the published patent application explains. “When the third-party game notifies the game platform of the improperly awarded achievement, the game platform may award the player an item in response to the achievement, thereby rewarding the cheating behavior.”
The tech that Microsoft have filed a patent for will basically sift through data on interactions between games, and gaming platforms such as Xbox Live, to track things such as achievements and gamer scores, and player rankings. The AI will analyze this data, and look for any cheating based on abnormal activities. An AI would, of course, make sense for this kind of work, since the data in question would simply be too huge. While manual analysis of such data is possible, the same work being done by an AI would be more uniform, more efficient, and quicker.
Microsoft have also proposed a “goals management module” to support this cheating detection AI. The goals management module is essentially going to make sure that players only get rewards such as achievements once they’ve met the requirements for such awards, ensuring that they don’t unlock the same through underhanded methods. “The goals management module confirms that the player’s information meets the relevant policy, and then updates the goals information in the user accounts database to show that the goal has been met (and to award any items for meeting the goal),” the patent explains. “Goals information also may contain information regarding the meeting of goals in other platform activities.”
The cheating detection tech will then compare the data gathered through the goals management module with data that comprises of gameplay that doesn’t involve cheating, and after cross checking this, it will detect if there are any players whose accounts should be flagged for irregular activities or cheating in games. If an account is flagged, what action is taken against the player will depend on their history. First time offenders might receive a warning e-mails, while repeat offenders might automatically get banned.
To say the least, it sounds interesting. Self-learning AI sounds like a good fit for something like this, especially when you take into account just how large the entire data pool in question would be. It remains to be seen whether Microsoft will be able to fine-tune this tech enough to be able to bring it into large-scale practice, and when (or if) this will become an actual thing, rather than just an idea on paper, which is what it is right now.