Xbox Network Blocks 4.7 Million Pieces of Inappropriate Content

Read this post in :

Microsoft has recently released the first half of the 2023 Transparency Report, including the 6 month period from January to June. This report describes the type of approach that Microsoft is taking in securing the Xbox service as a whole and keeping it safe for everyone. As well as showing via graphs the total numbers of proactive as well as reactive enforcement of the policy. Which includes not just security related cases like phishing and account tampering, but also piracy, fraud, cheating and vulgar content as well.

AI tools Integration

This year’s report matches the current state of the industry, moving toward an AI enhanced service for better or for worse. Named Community Sift, it combines both AI moderation tools along with human-driven ones, to potentially scale to a more manageable format for the  workers. With support of up to 22 languages, including emojis and the culture nuance of each language. As well as using the Turing Bletchley v3, being able scan images generated by the user base and block any inappropriate content, which helped blocking 4.7 MIllion of harmful content.

H1 2023 vs 2022

This half of the year is obviously compared to last year’s report. With Microsoft showing via data and graphs the situation, as well as acknowledging and comparing them with the previous year. Microsoft tends to clarify the difference between Proactive and Reactive enforcement. The first stands for preventive action done before it can cause any harm, like text videos and images. While reactive is done via reports from users themselves.

This year sees a far more Proactive Enforcement, with an increase of 2X over last year, with about 95% revolving around tampered accounts generally aimed at cheating and causing disruption for other players. Not to mention a 184% increase in proactive enforcement of vulgar content. Still Reactive Enforcement was about on par with last year’s report, with a slight increase in the overall comment type report, such as offensive messages left or sent.

Takeaways

This year sees a staggering increase in enforcement over cheating and fake accounts, with 16,29 Millions over last year 4,33 Millions. a small increase in bullying and harassment. Not to mention appeal cases almost doubled from 2022’s 151K to 268K. On one hand it seems that AI integration helped weed out a lot of fraudulent accounts because they could cause disruption;  but on the other hand it had increased over 4 times meaning that the overall activity has also increased and not just being detected easier.

Moreover the appeal cases have dramatically increased, with far more Non-Reinstatement, while Reinstatement has also decreased percentage wise. AI integration seems to have increased the overall prevention of fraudulent and harmful activities, it’s not undeniable that the need for Human moderation is absolutely necessary. The scary thought is that in the long run Microsoft may switch to a fully AI moderation tool, just like many of the other major companies are hoping to, just to cut corners and grow their annual revenue, potentially leaving AI moderation fighting AI generated phishing and frauds. Potentially resulting in a much more strict and heavily censored platform for regular harmless users to use.

PURCHASE XBOX GAME PASS AND WINDOWS PRODUCTS FOR CHEAP ON ALLKEYSHOP

Allkeyshop is the best place to buy the best games at the lowest prices, and we even have a browser extension that helps you find the best discount codes, coupons, and prices instantly.

Download the Allkeyshop Browser Extension Free – SAVE MONEY

For all the latest video game news, trailers, and best deals, make sure to bookmark us.

You can find all the best and cheapest online deals on CD keys, game codes, gift cards, and antivirus software from the verified CD key sellers on our store pages.

By  whizzerlow

More relevant articles