Interesting. I’m sure he has not identified everything, but this is a good start.
https://www.schneier.com/blog/archives/2023/11/ten-ways-ai-will-change-democracy.html
Ten Ways AI Will Change Democracy
Artificial intelligence will change so many aspects of society, largely in ways that we cannot conceive of yet. Democracy, and the systems of governance that surround it, will be no exception. In this short essay, I want to move beyond the “AI-generated disinformation” trope and speculate on some of the ways AI will change how democracy functions—in both large and small ways.
… Some items on my list are still speculative, but none require science-fictional levels of technological advance. And we can see the first stages of many of them today. When reading about the successes and failures of AI systems, it’s important to differentiate between the fundamental limitations of AI as a technology, and the practical limitations of AI systems in the fall of 2023. Advances are happening quickly, and the impossible is becoming the routine. We don’t know how long this will continue, but my bet is on continued major technological advances in the coming years. Which means it’s going to be a wild ride.
A question that really needs an answer.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4623126
Criminal Liability of Artificial Intelligence
Artificial intelligence is a new and extremely quickly developing technology, which is expected, and maybe even feared to bring enormous changes in every aspect of our society. Even though this technology is still comparatively underdeveloped, we already hand over a multitude of everyday-tasks. As for now AI is mostly used to take over tasks, which are often perceived as “annoying” or highly time consuming. Therefore, it shall enhance productivity in first place. It is expected to do many of the tasks even better than human beings. At least in future. Some of these tasks, such as autonomous driving are quite dangerous, bearing the potential to infringe peoples protected rights, and even cause physical harm and death to human beings. Obviously, such technology needs a solid and reliable legal basis, especially in terms of liability, if the inevitable happens and the technology causes events that were not intended to happen. However, a well-developed set of rules should not only concern private law. Especially when such technology causes harm or even death to human beings, the question of a criminal deed arises, in a sense of criminal negligence for example. Future criminal law must be prepared and probably adjusted effectively tackle any questions concerning criminal liability of artificial intelligence.
Are we evolving toward an AI lawyer?
http://192.248.104.6/handle/345/6771
Impact of Artificial Intelligence on Legal Practice in Sri Lanka
Artificial Intelligence (AI) a machine-based system used to ease the human workload, has been popular globally and its influence can be seen even in developing countries like Sri Lanka. Although it has dominated areas such as machine problem detecting, calculating and speech recognition, it is questionable whether this sophisticated technology can address the traditional roles of legal practice. The research aims to explore the positive and negative influence of AI in the legal field while determining the degree to which this technology should be incorporated into the legal sector in Sri Lanka. The research was carried out as a literature survey with a comparative analysis of other jurisdictions. Currently, many countries including the USA have used AI-based tools such as LawGeex, Ross Intelligence, eBrevia and Leverton in legal practice due to their efficiency, accuracy and ease of use. Findings revealed that AI can be used even in Sri Lanka for legal research, preliminary legal drafting and codification of law. But according to the prevailing economic and social background of Sri Lanka, it will be discriminatory to totally rely on an AI-induced legal system since it may create barriers to equal access to legal support for the common masses. Also, excessive dependency on AI will be a barrier to innovative legal actions such as public interest litigation since it would not assess the humanitarian aspect. Hence, it is concluded that AI should be used in Sri Lankan legal practice with limitations.
Thoughtful. Something for Con-Law at last!
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4626235
AI Outputs and the Limited Reach of the First Amendment
Not all communications are “constitutional speech” - determining whether machine-generated outputs qualify for First Amendment protection requires some work. In order to do so, we first explore aspects of both linguistic and communication theories, and then under what circumstances communication can become First Amendment speech.
We reach the bounds of the First Amendment from two directions. Working from a linguistic definition of speech, we capture non-linguistic forms of protected speech. Using communication theory, we reach a divide between human-sender communication and non-human-sender communication. Together these approaches support the location of a constitutional frontier. Within we find all instances of recognized First Amendment effectiveness. Outputs of non-human autonomous senders (e.g. AI) are outside and constitute an unexamined case.
“Speech” under the First Amendment requires both a human sender and a human receiver. Concededly many AI outputs will be speech – due to the human factor in the mix. But just because a human programmed the AI, or set its goals, does not mean the AI’s output is substantially the human’s message. Nor does the fact that a human receives the output, for listener’s First Amendment rights arise only where actual speech occurs. Thus, we resist the claim that all AI outputs are necessarily speech. Indeed, most AI outputs are not speech.
For those who raise objection to the challenge we pose – determining which AI outputs are speech and which are not – we respectfully note that there will be additional Constitutional work to be done. We are confident that our courts will be up to this challenge.
Whether AI outputs are First Amendment speech has profound implications. If they are, then state and federal regulation is severely hobbled, limited to the few categories of speech that have been excluded by the Supreme Court from strong constitutional protection.
With limited exception, neither the sponsors/developers of AI, the AI itself, nor the end users have rights under the First Amendment in the machine’s output. We express no opinion on other rights they may have or on what types of regulations state and federal governments should adopt. Only that they may constitutionally do so.
(Related) They may have put a finger on the problem. AI output is based on the data it has scanned.
https://ojs.journalsdg.org/jlss/article/view/1965
The Impact of Developments in Artificial Intelligence on Copyright and other Intellectual Property Laws
Objective: The objective of this study is to investigate the impact of AI breakthroughs on copyright and challenges faced by intellectual property legal protection systems. Specifically, the study aims to analyze the implications of AI-generated works in the context of copyright law in Indonesia.
… Result: The research findings reveal that according to Law Number 28 of 2014 in Indonesia, AI-generated works do not meet the originality standards required for copyright protection.
This interests me because of the years I spent auditing computer systems.
https://link.springer.com/article/10.1007/s44206-023-00074-y
Auditing of AI: Legal, Ethical and Technical Approaches
AI auditing is a rapidly growing field of research and practice. This review article, which doubles as an editorial to Digital Society’s topical collection on ‘Auditing of AI’, provides an overview of previous work in the field. Three key points emerge from the review. First, contemporary attempts to audit AI systems have much to learn from how audits have historically been structured and conducted in areas like financial accounting, safety engineering and the social sciences. Second, both policymakers and technology providers have an interest in promoting auditing as an AI governance mechanism. Academic researchers can thus fill an important role by studying the feasibility and effectiveness of different AI auditing procedures. Third, AI auditing is an inherently multidisciplinary undertaking, to which substantial contributions have been made by computer scientists and engineers as well as social scientists, philosophers, legal scholars and industry practitioners. Reflecting this diversity of perspectives, different approaches to AI auditing have different affordances and constraints. Specifically, a distinction can be made between technology-oriented audits, which focus on the properties and capabilities of AI systems, and process-oriented audits, which focus on technology providers’ governance structures and quality management systems. The next step in the evolution of auditing as an AI governance mechanism, this article concludes, should be the interlinking of these available—and complementary—approaches into structured and holistic procedures to audit not only how AI systems are designed and used but also how they impact users, societies and the natural environment in applied settings over time.
(Related) You mean I can generate my own version of the evidence!
An Overview of Deepfake Technologies: from Creation to Detection in Forensics
Advancements in Artificial Intelligence (AI) techniques have given rise to significant challenges in the field of Multimedia Forensics, particularly with the emergence of the Deepfake phenomenon. Deepfakes are images, video and audio generated or altered by powerful generative models such as Generative Adversarial Networks (GANs) [5] and Diffusion Models (DMs) [12]. While GANs have long been recognized for their ability to generate high-quality images, DMs offer distinct advantages, providing better control over the generative process and the ability to create images with a wide range of styles and content [2]. In fact, DMs have shown the potential to produce even more realistic images than GANs. The AI-generated contents span diverse domains, including films, photography, video games, and virtual reality productions. A major concern of the Deepfake phenomenon is the application on important people such as politicians and celebrities to spread misinformation. However, the most alarming aspect is the misuse of GANs and DMs to create pornographic Deepfakes, posing a serious security threat. Notably, a staggering 96% of Deepfakes available on the internet fall into this pornographic category. The malicious use of Deepfakes extends to issues such as misinformation, cyberbullying, and privacy violation. In addition, Deepfakes have been applied in the fields of art and entertainment, sparking ethical discussions about the limits of creativity and authenticity. To counteract the illicit use of this powerful technology, novel forensic detection techniques are required to identify whether multimedia data has been manipulated or altered using GANs and DMs. Regarding image deepfake detection methods in the state of the art, the primary focus lies in binary detection, distinguishing between Real and AI-generated images [14, 16]. Notably, some methods in the state of the art have already demonstrated the ability to effectively differentiate between various GAN architectures [4, 7, 6, 15] and several DM engines [13, 1, 9]. These researches showed that generative models leave unique fingerprints in the generated multimedia data, which can be used not only to identify Deepfakes, but also to recognize the specific architecture used during the creation process [11]. This can be extremely important in forensics in order to reconstruct the history of the multimedia data under analysis (forensic ballistics) [8]. In order to create increasingly sophisticated deepfakes detection solutions, several challenges have been proposed by the scientific community such as the Deepfake Detection Challenge (DFDC) [3] and the Face Deepfake Detection Challenge [10]. The latter has also launched a new challenge among researchers in the field: reconstructing the original image from deepfakes; a task that can be extremely important in forensics.
No comments:
Post a Comment