AI is a double-edged sword, bringing great convenience to people while also significantly increasing the level of fraud and false propaganda. Evil regimes like the Chinese Communist Party are attempting to use AI deepfake technology to mislead public opinion and disrupt the international community.
AI technology is rapidly advancing, allowing more users to expand and enhance their work through open AI code, application programming interfaces (APIs), etc. However, regulations and oversight related to AI have not kept pace, enabling evil regimes to use “deepfake technology” to spread false propaganda and disrupt public opinion.
Deepfake technology involves replacing faces or altering speech content in videos, with advanced deepfake technology reaching a point where it is highly realistic.
Recently, a video surfaced on China’s Douyin platform featuring former U.S. President Trump and current Vice President Harris singing a Chinese Communist Party propaganda song “I Love You, China” with Chinese accents.
Independent writer Zhu Ge Mingyang believes that, “Although everyone knows this is a spoof, due to its novelty and entertainment value, it attracts a large number of viewers to watch such videos, inadvertently brainwashing the Chinese people.”
Moreover, the Chinese Communist Party’s social platforms use unauthorized images of domestic and foreign celebrities, influencers to promote certain concepts, political ideologies, or provide free advertising.
For instance, a 20-year-old Ukrainian girl Olga Loiek was shocked to find her image appearing on Chinese social platforms like Xiaohongshu and Bilibili, despite never using these platforms and only sharing her thoughts and life on YouTube and Instagram.
She was further surprised and angered to discover 35 “doppelganger” accounts bearing her resemblance on Chinese social media platforms, falsely claiming she has lived in China as a “Russian” for many years. These accounts propagate content similar to Chinese Communist Party propaganda, often expressing desires to marry Chinese individuals and promoting a strong alliance between China and Russia, gaining significant traction.
After starting her YouTube channel last year, Loiek’s face was unlawfully manipulated using deepfake technology, predominantly originating from Shenzhen, China-based tech company “Shi Yun Technology”, now known as HeyGen, headquartered in Los Angeles. The company’s founders previously received substantial investments from major Chinese venture capitalists.
In addition to Loiek’s case, in February last year, the social media company Graphika released a report exposing the relationship between a network media outlet named “Wolf News” and the Chinese Communist Party. The report revealed that “Wolf News” utilized deepfake technology to create two fake foreign anchors broadcasting fake news 24/7, predominantly promoting and denigrating European and American societies for the Chinese Communist Party.
Founder of Taiwan AI Labs, Du Yijin, told Epoch Times, “In the past, videos or live broadcasts generated using deepfake technology had many loopholes, but now these flaws are decreasing, even reaching a point where they are hard to detect. Therefore, in the future, discerning the authenticity of such videos and live broadcasts will be a major challenge for people.”
The use of AI to generate large numbers of fake accounts and articles is pushing the “spamouflage” in information warfare and technology to deeper and broader levels, making such information even harder to detect.
Du Yijin pointed out that the Chinese internet navy, following two assassination attempts on Trump in July and September, hijacked users’ accounts on platforms like YouTube, X, and Facebook, or created numerous fake accounts using AI, posing as supporters or opponents of Trump. These fake accounts would first repost articles pretending to be real users, then later post extreme comments to incite tension between Trump supporters and opponents, aiming to divide American citizens.
In mid-last month, OpenAI published a report detailing how Iran used ChatGPT to generate a vast amount of media comments, articles, and messages, distributed on platforms like Instagram and X.
The report revealed they first had AI write articles on U.S. politics and global events, published on five websites simultaneously acting as progressive and conservative news outlets, then used AI to generate short comments in English and Spanish, posted on social media as real users to spread false information.
The findings disclosed in this report were similar to OpenAI’s late May revelation of similar activities conducted by China, Russia, Iran, and a privately-owned Israeli company. They found activities by Iran’s “International Virtual Media Alliance”, China’s “Spamouflage” organization, utilizing OpenAI’s AI to analyze public social media activities and disseminate false messages in multiple languages on major social platforms, aiming to manipulate public opinion and influence geopolitics.
Graphika also revealed in 2019 and 2021 how China utilized “spamouflage” to sow division in the international community and spread massive false information.
The company stated that China first used AI or internet navy to create numerous fake accounts, flooding social media platforms with “spam posts” to disguise their true intentions and evade detection. They would wait for significant global events or anti-Chinese sentiments to emerge before these accounts began praising China or posting a slew of negative comments aiming to discredit anti-Communist governments and organizations.
Du Yijin expressed, “Previously, our lab also discovered numerous AI accounts on social platforms, often during Taiwan and U.S. presidential elections, masquerading as supporters of different candidates and frequently employing extreme rhetoric to provoke hate between supporters, making it difficult for them to engage in civilized discussions on real issues, further escalating societal divisions and conflicts.”