Stay Informed with Artificial Intelligence Updates
TikTok, the popular social media platform owned by China’s ByteDance, has taken steps to address growing concerns about the spread of online disinformation and deepfakes. In a first for any social media platform, TikTok has announced that it will automatically label content generated through artificial intelligence (AI). This move comes as generative AI technology advances and raises questions about the authenticity of digital media.
Currently, online platforms like Facebook and TikTok require users to disclose if their content is created using AI software. However, TikTok has gone a step further by developing its own features to identify and label videos that are AI-generated. This includes content produced using Adobe’s Firefly tool and TikTok’s own AI image generators, as well as OpenAI’s Dall-E.
Adam Presser, TikTok’s head of operations and trust and safety, emphasized the importance of distinguishing between human-generated and AI-generated content. He acknowledged the rise in harmful AI-generated content and noted that authenticity is a key element that has contributed to TikTok’s vibrant and joyful community.
While various social media platforms, including TikTok, Meta, X, and YouTube, have explored integrating generative AI into their platforms, they have faced criticism for allowing users to be flooded with low-quality AI-generated spam content.
These platforms face additional challenges during major elections as they are pressured to introduce guardrails against misleading deepfakes, covert influence operations, and ensure fair content moderation while remaining non-partisan. This comes in the wake of TikTok and its parent company, ByteDance, filing a lawsuit against the US government to challenge a law aimed at forcing a sale or ban of the app due to concerns about potential disinformation and propaganda.
To tackle these concerns, TikTok has joined a coalition led by Adobe that focuses on incorporating content credentials into AI-generated products. This coalition aims to embed digital fingerprints into multimedia AI content, including information about when, where, and by whom the material was generated. TikTok plans to leverage these indicators to automatically flag content created using AI products.
OpenAI, a leading AI research organization, has also joined the coalition and intends to embed fingerprinting technology into all images generated by their image model, Dall-E 3. Additionally, OpenAI plans to embed these indicators into its video-generating model, Sora, once it is widely released.
Recognizing the significance of this development, large technology companies such as Google, Microsoft, and Sony are also exploring the integration of similar technologies into their AI tools. Meta, formerly known as Facebook, has announced its intentions to stamp AI-generated content with a “Made by AI label” by detecting hidden markers inserted by organizations like Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock. The company is also actively developing deepfake detection classifiers.
It is essential to acknowledge that bad actors and sophisticated disinformation groups may exploit open-source AI generation tools to create deepfakes, making them harder to trace and unlikely to be flagged by digital fingerprinting and watermarking. While tech companies argue that their current efforts are a step towards addressing this problem, the implications and potential future trends remain significant.
Looking forward, it is crucial for the industry to prioritize transparency and provide authentic, trustworthy conversations in a digital world where manipulated content is pervasive. Data and technology companies, like Adobe, play a vital role in enabling these interactions. Dana Rao, general counsel and chief trust officer at Adobe, highlights the premise of their solution as a means to foster transparency and genuine communication with the public, considering the manipulative nature of digital content today.
In conclusion, TikTok’s decision to automatically label AI-generated content is commendable given the rising concerns surrounding deepfakes and disinformation. It serves as a reminder that authenticity and transparency are pivotal in nurturing vibrant online communities. As technology continues to evolve, industry leaders must forge alliances to combat the challenges posed by AI-generated content.