Artificial intelligence’s capacity to generate text has surpassed mere summarization and translation, entering the domain of “persona mimicry” — replicating specific individuals’ writing styles and modes of reasoning. This advance has thrown into sharp relief the ethical and legal challenges of “AI ghostwriting” and the manipulation of public opinion through fabricated digital identities.

A 2025 study found that LLM use reduced critical thinking scores, while research published by Cornell University in 2025 indicated that using large language models can lead to cognitive atrophy and reduced brain elasticity, as neural networks related to memory become underused. According to Stanford University’s 2025 AI Index, global private investment in generative AI reached US$33.9 billion in 2024 — over 20% of all AI-related private investment, up 8.5× from 2022.

1.1 Cyber Pseudepigraphy and the Hollowing of Authorship

When AI communicates while pretending to be human, it violates individual autonomy and corrodes social trust. In education and research, AI-assisted dishonesty has emerged as a novel form distinct from traditional plagiarism — termed cyber pseudepigraphy: having AI generate content without engaging one’s own thinking, then presenting the output as evidence of one’s intellectual capacity. Whereas traditional plagiarism appropriates another person’s work, AI generation produces text that “nobody wrote — including the person presenting it,” hollowing out the very concept of authorship.

1.2 The Digital Twin: Legal Status and the Risk of Being Propertized

Jennifer Rothman, Nicholas F. Gallicchio Professor of Law at the University of Pennsylvania and an international authority on the right of publicity, has recently focused her scholarship on “the ways intellectual property law is employed to turn people into a form of property.” As AI enables the creation of human digital twins (HDTs) that fully replicate an individual’s appearance, voice, and personality, she warns that Congress is on the verge of a serious mistake.

The European Parliament Research Service’s December 2025 briefing warns that generative AI has dramatically accelerated foreign information manipulation and interference (FIMI). Since Russia’s full-scale invasion of Ukraine, pro-Russian disinformation campaigns powered by generative AI have surged. In March 2022, a deepfake video of President Zelensky apparently ordering soldiers to surrender spread across Twitter, Facebook, YouTube, and VKontakte.

2.1 Operation Overload and the Industrialization of Impersonation

Between January and March 2025, the Institute for Strategic Dialogue (ISD) found that a Russia-aligned campaign, “Operation Overload,” impersonated over 80 different organizations — exploiting their logos, fonts, and staff voices. One video used the Wall Street Journal’s branding to falsely claim that USAID funded “LGBTQ+ values among kids in Ukraine.” The technique weaponizes the credibility of established institutions themselves.

As AI evolves into “agentic” systems capable of autonomously executing code, signing contracts, and completing transactions, regulatory focus is shifting from static content generation to dynamic decision-making and action. Reinterpretation of agency law is already underway to determine whether liability for a harmful AI-negotiated contract falls on the user or the developer.