moderation
Latest
Meta reportedly won't make its AI advertising tools available to political marketers
Reuters reports that Meta will specifically not make its AI advertising tools available to political marketers ahead of the presidential election cycle.
Xbox’s reporting system for abusive-voice chat arrives this week
Microsoft is set to launch its abusive voice reporting feature for Xbox consoles. Announced in July, it lets gamers submit inappropriate remarks heard while playing multiplayer titles. The system captures a 60-second clip saved to the console; you then have 24 hours to complete the report. The feature arrives this week in the September update for Xbox Series X/S and Xbox One. However, it’s initially limited to the “select English-language markets” of the US, UK, Canada, Ireland, Australia and New Zealand.
ChatGPT is easily exploited for political messaging despite OpenAI's policies
ChatGPT was supposed to be inoculated against political misinformation in March. Doesn't look like that actually happened.
Twitch streamers can soon block banned accounts from tuning in
Twitch announced this week that an upcoming change will allow streamers to block banned users from tuning into their streams. “You can choose to have your banned chatters no longer be able to watch the stream,” Senior Product Manager Trevor Fisher revealed on Twitch’s Patch Notes podcast (via TechCrunch), stressing that the feature won’t be enabled by default. The new blocking feature will roll out in the next few weeks.
Anthropic explains how Claude's AI constitution protects it against adversarial inputs
Anthropic explains how binding its Claude AI to a set of guiding principles will lead to better outputs and prevent racist meltdowns.
Bipartisan bill would require that social networks have 'clear' content policies
Senators have introduced a bill that would make social networks provide clearer content moderation policies.
Google is making free anti-terrorism moderation tools for smaller websites
Google is helping smaller websites fight terrorist content by making a free moderation tool.
Twitch halts paid stream boosts after viewers abuse them to push porn
The new program lets viewers boost their favorite streams when they purchase subscriptions and bits.
GGWP is an AI system that tracks and fights in-game toxicity
AAA studios won’t fix the problem, so esports pro Dennis Fong is stepping up.
Instagram Live creators can now bring in moderators to handle trolls
The feature could make broadcasts a more positive experience for almost everyone.
Twitter reportedly knew Spaces could be misused due to a lack of moderation
Company executives forged ahead with the audio chat feature despite warnings, according to 'The Washington Post.'
Twitch offers slightly more information about suspensions
But won't just come out and say why.
Senate Republicans vote to subpoena Facebook and Twitter CEOs
Senate Republicans voted to subpoena Facebook and Twitter CEOs to testify about their alleged blocking of a controversial New York Post article.
Sony tries to clear up confusion over voice chat recording on PS5
There is no opt-out of this — if you use voice chat on PS5, someone in your chat can clip your audio and submit it for review by the moderation team.
Reddit has banned nearly 7,000 hateful subreddits since June 29th
Reddit has banned nearly 7,000 hateful subreddits since its new content policies went into effect.
Pinterest's moderation doesn't catch some abusive and false material
Pinterest's emphasis on hiding harmful content over removing it has left some abusive and false material visible.
Twitch forms review board to strengthen moderation policies
Twitch’s reputation needs a spit-shine. The Safety Advisory Council is an attempt to do just that.
Facebook will pay content moderators $52 million in PTSD settlement
Each of the 11,250 plaintiffs will receive at least $1,000.
YouTube will temporarily increase automated content moderation
YouTube will rely more on machine learning and less on human reviewers during the coronavirus outbreak. Normally, algorithms detect potentially harmful content and send it to human reviewers for assessment. But these are not normal times, and in an effort to reduce the need for employees and contractors to come into an office, YouTube will allow its automated system to remove some content without human review.
TikTok will stop using China-based moderators to screen foreign content
TikTok has already taken steps to reassure the world that the Chinese government doesn't control its app overseas, including the use of non-Chinese moderators for the US and plans for a transparency center. However, it's taking things one step further. The social media company said it will stop using China-based moderators to screen content in any other country, and that more than 100 moderators will have to either find other jobs inside parent company Bytedance or leave. Teams local to given areas should take over within a few weeks, TikTok said.
