Spectrum Labs Launches Content Moderation for Generative AI

Applications of AI


AI Battle: Spectrum Labs Combats AI-Generated Hate Speech, Radicalization, and Grooming with Industry-First AI-Powered Content Moderation

Miami, April 6, 2023 /PR Newswire/ — spectrum labText Analysis AI, a leader in providing tools to scale content moderation for games, apps and online platforms, today announced the world’s first AI content detection and prevention of harmful and toxic behaviors produced by generative AI. Announcing the launch of our moderation solution. With the rise of generative AI such as ChatGPT, Dall-E, Bard, and Stable Diffusion, automated content creation is being used to create racist imagery, hate speech, radicalization, spam, fraud, grooming, and harassment. It can now spread rapidly and continuously. Large-scale, short-term investment by bad actors intent on exploiting new technology.

To start addressing this problem, Spectrum Labs has developed a first-of-its-kind moderation tool for generative AI content. This tool helps the platform auto-protect the community from this highly scalable hostile content.

“Even before generative AI, platforms sifted through tons of user-generated online content every day and struggled to identify and remove hateful, illegal and predatory content. Gang recruiter, your job just got a whole lot easier.” Justin Davis, CEO of Spectrum Labs. “Fortunately, our existing contextual AI content moderation tools are built to detect intent, not just lists of keywords and specific phrases that generative AI can easily avoid, so this new It can be adapted to deal with the flood of content.”

Generative AI is designed to create plausible variations of human speech, so traditional keyword-based moderation tools may not be able to identify content where certain racist words or phrases are not used. It is not possible to detect whether the intent of is hateful. (e.g., a child’s story about why one race is better than another without the racial slur). Similarly, existing context models that can detect sexual, threatening, or toxic content, but not positive behaviors such as encouragement, approval, and intimacy, suggest that content is intended to be informative, supportive, or reassuring. Edit generative AI responses on sensitive topics, even those that have (For example, a user struggling with sexual abuse is seeking help finding resources for her psychological support).

Even for image-based generated AI such as Dall-E, automatic detection and editing of toxic human-generated prompts can prevent the creation of libraries of new AI-generated image and video content such as hate, intimidation and extremism. . Real-time latency that makes the user experience of generative AI feel like magic.

Future uses of Generative AI’s multi-layer real-time AI moderation include detecting piracy, detecting bias in AI-generated content to filter and eliminate biased and problematic training data sources, and people may include better analysis of the types of content required by How to make and how to use. But now, the company is focused on rapidly delivering a foundational set of tools to help protect its users and platform from a potential tsunami of toxic content.

“At Spectrum Labs, we are on a mission to make the internet a safer place for all. It’s an honor to support you. Together, we’re building a safer digital world, one post at a time,” added Davis.

For more information on Spectrum Labs’ AI-generated content moderation, please visit: https://www.spectrumlabsai.com/generative-ai-content-moderation

About Spectrum Labs
As the leader in text analytics AI, Spectrum Labs harnesses the power of data, AI, and community to build a better internet, keep billions of users safe online, and deliver greater value to businesses, consumers, and society. A place to be. Using contextual AI and its latest innovations, Healthy Behaviors AI, games and apps eliminate toxic users and content at scale while building positive user interactions into their designs and increasing retention through a more positive user experience. , revenue, and engagement.Let’s build a better internet – visit SpectrumLabsAI.com you can know more.

media contact
[email protected]

Source Spectrum Labs



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *