Research On AI Deception In Deepfakes And Its Regulation

The ushering in of the imagery era has led to a transformation in the manner in which the public formulates and retrieves memories,with visual misrepresentations more likely to engender erroneous cognition compared to verbal content.The rapid progress of digital technology,coupled with the increased access to multimedia information,has facilitated the manipulation and distortion of digital image authenticity.Deepfake technology,specifically,anchored in deep learning and artificial intelligence,is fundamentally altering the dynamics of social media and the fabric of human existence.The focus of this study lies on the "stitched reality" brought about by deepfake technology.Employing the "deception theory" and using methodologies such as textual analysis and qualitative comparison,the study delves into the global risks and challenges posed by this sophisticated form of "deception".It is posited by the research that the concept of "deception" will be redefined by deepfakes in terms of intent,content,and action,rendering each aspect more intricate and intelligent.This evolved form of deception,termed as "intelligent deception," arises from the collective action of various entities.Particularly,the spread of deepfakes is facilitated by the advancements in image synthesis and editing technologies,while the dissemination of false information is propelled by political manipulation and economic motives.The open-source distribution of AI technology speeds up the spread of deepfakes,and the audience’s involvement and experiences guarantee their sustained relevance.All these elements contribute to the commercialization of deepfakes,thereby posing a risk to the definition of reality.In essence,the evolution of deepfake technology can be bifurcated into two stages-the popularity phase and the commercial phase.During the popularity phase,the deceptive,automated,manipulative,and multi-modal attributes of deepfakes became progressively conspicuous.With the onset of the commercial phase,the risks associated with deepfakes expanded further.The control over public attention has transitioned from credible media organizations to the creators of deepfakes.Malicious deepfake information not only interrupts the truth mechanism but also gives rise to legal issues such as exploitation via pornography,privacy infringements,and can even influence the decision-making of political leaders and significant public figures,thereby affecting global geopolitics and national security.It is therefore of utmost importance to digitally regulate deepfakes.In terms of digital regulatory provisions,it is crucial for government agencies to continually strengthen pertinent laws,regulations,and supervisory mechanisms,actively advocate for the global governance of deepfakes,and elevate users’ digital safety awareness.As for digital regulatory mechanisms,it is incumbent upon online platforms to identify and filter deepfake content to avoid risks of forgery and technological dominance.Regarding digital regulatory technology,considering the stealth and subtlety of deepfake technology,custom strategies should be developed to address different forms of "fusion".In terms of digital regulation models,adopting mature foreign practices,aligning them with China’s unique circumstances and cultural background,and establishing a cooperative governance mechanism involving multiple stakeholders is a viable option to enhance users’ digital literacy education.