AI and You Copyright Exemption for Nonhuman Creations, AI ‘Boss From Hell,’ Porn Deepfakes

AI and You Copyright Exemption for Nonhuman Creations, AI 'Boss From Hell,' Porn Deepfakes

The Copyright Battle between AI and Human Creativity

In a landmark ruling by US District Judge Beryl Howell, it was determined that art created by AI does not qualify for copyright protection. The ruling upheld the decision by the US Copyright Office that non-human entities are not entitled to copyright protection. Judge Howell emphasized that human authorship is a crucial requirement for copyright protection.

This ruling has implications not only for art created by AI, but also for works generated with other technological tools, such as cameras. While cameras produce mechanical reproductions of scenes, they do so based on the initial creative decisions made by humans, such as lighting, arrangements, and subject positioning. This distinction emphasizes the importance of human creativity in copyrightability.

The debate around copyright laws and AI becomes even more complex in the context of generative AI tools used to train large language models. These tools scrape the internet for copyrighted material without permission or compensation, leading to legal disputes between authors, artists, and the creators of chatbots powered by AI. Cases involving these disputes are ongoing, and their outcomes will shape the future of copyright laws in relation to AI technology.

The issue of using AI to replace unionized labor in creative industries is also a growing concern. As Hollywood creatives go on strike, fearing that studios may employ AI to author scripts and generate other forms of creative work, the potential for AI to encroach on human creativity becomes a pressing issue. However, the Copyright Office has acknowledged that in certain instances, AI-assisted materials may qualify for copyright protection if the human element in their creation is sufficiently creative.

The interaction between copyright laws and AI presents a challenging landscape for developers of generative AI chatbots. As the use cases for AI technology evolve, navigating the legal implications of copyright infringement becomes increasingly important. It is crucial for AI creators to consider the complex interplay between AI-generated content and copyright laws.

Questioning the Reliability of AI Search Results

Google’s experimental genAI search service recently encountered some troubling issues. The AI-powered search engine displayed results that included justifications for slavery, genocide, and the benefits of banning books. It even provided recipes for poisonous mushrooms. These problematic search results were brought to light by a senior director of Search Engine Optimization, raising concerns about the AI’s lack of understanding of context and potential to deliver offensive information.

Google addressed the situation, stating that the search experience was limited to users who had opted into the experimental program. However, this incident highlights the challenges AI systems face in providing accurate and contextually appropriate search results.

Professional Headshots and Deepfakes

Busy professionals are turning to AI systems to enhance their appearance in professional headshots. A myriad of apps and online services allow individuals to select a desired style or aesthetic and have their photos transformed accordingly. Users have reported that no one noticed AI involvement in the retouching process. However, glitches can occur, such as the addition of extra body parts or inconsistent enhancement of features like teeth and ears. The AI’s training sets may also introduce biases that alter complexion, skin tones, and eye color, as demonstrated by a postgraduate student whose eye color was changed from brown to blue.

Meanwhile, in Hong Kong, authorities arrested a group of individuals who used AI-based deepfake technology to change their faces and deceive banks into granting loans. Criminals would modify their appearance to match stolen identity cards and use these false credentials to apply for loans. The syndicate also utilized stolen identities to register for multiple SIM cards, which they used for phishing attempts to acquire credit card details and personal information.

The Rise of Pornographic Deepfakes

The creation of non-consensual pornographic deepfakes, which involve using AI to fabricate explicit videos featuring the faces of real women, has been surging. Research indicates that the number of deepfake videos has increased ninefold since 2019, with nearly 150,000 videos across 30 different sites. This trend poses significant challenges, as there are currently no federal laws in the United States criminalizing the creation or sharing of non-consensual deepfake porn. Experts argue that the responsibility lies with tech companies to self-regulate and implement measures to protect individuals’ rights and identities.

AI in the Workplace: Potential Challenges

While AI tools like ChatGPT offer impressive capabilities for generating human-like text, concerns have been raised regarding their potential impact on the workplace. University of Oxford Professor Michael Wooldridge, who specializes in computer science, warns that AI could become an intrusive and controlling supervisor, monitoring employees’ emails, providing continuous feedback, and even making decisions about job termination. While the development of powerful AI tools like ChatGPT is indeed remarkable, understanding the risks associated with AI technologies is crucial.

For instance, AI systems can analyze social media feeds to discern users’ political leanings and manipulate them through disinformation campaigns. Hence, encouraging skepticism, promoting transparency, ensuring accountability, and addressing the limitations and flaws of AI chatbots are vital steps in evaluating, managing, and mitigating the risks posed by AI in the workplace.

Beyond the Turing Test: Assessing Consciousness

The infamous Turing Test, proposed by Alan Turing, was designed to assess a machine’s ability to exhibit intelligence indistinguishable from that of a human through text-based conversations. While some argue that modern chatbots like ChatGPT have passed the Turing Test due to their conversational abilities, others contend that true understanding and consciousness are still beyond the reach of AI.

In an effort to delve further into the question of consciousness, a group of experts in neuroscience, philosophy, and computer science have developed a checklist of criteria to determine whether a system exhibits a high chance of being conscious. This framework aims to define and test for consciousness using various theoretical approaches. While they admit that their framework is not yet complete, their evaluation of ChatGPT suggests that existing AI systems are not strong candidates for consciousness.

Conclusion

The copyright battle between AI and human creativity raises important questions about intellectual property and the role of AI in creative industries. As the use of generative AI technology expands, determining the boundaries of copyright protection becomes a complex challenge. In addition, the shortcomings and potential risks associated with AI in various domains, such as search engines, professional photography, and explicit content, require careful consideration and responsible use.

It is crucial for society to grapple with the ethical, legal, and societal implications of AI, ensuring that proper safeguards are in place to protect individuals’ rights and prevent the misuse of AI technologies. By fostering skepticism, promoting transparency, and embracing accountability, we can navigate the evolving landscape of AI and harness its potential while mitigating potential risks.