Category: AI Ethics

Posted on

From Machine Ethics to EU Law

In Article 50, “Transparency Obligations for Providers and Deployers of Certain AI Systems”, of the EU Artificial Intelligence Act, it is stated: “Providers shall ensure that AI systems intended to interact directly with natural persons are designed and developed in such a way that the natural persons concerned are informed that they are interacting with an AI system, unless this is obvious from the point of view of a natural person who is reasonably well-informed, observant and circumspect, taking into account the circumstances and the context of use.” On this subject, the European Parliament had already been advised ten years earlier by Oliver Bendel. In his lecture “Moral and Immoral Machines – Moralische und unmoralische Maschinen” in Brussels on September 8, 2016, he presented GOODBOT, a chatbot initiated by him in 2013 in the context of machine ethics, which featured several escalation levels while repeatedly making clear that it was merely a machine. At the Digital Europe Working Group Conference Robotics on November 8, 2017, Bendel also spoke online about related questions in machine ethics. In connection with a care robot, he raised the question: “Should the robot make clear that it’s just a machine?” The transparency obligations set out in Article 50 will enter into force on August 2, 2026.

Posted on

Towards Accessible Everyday Assistance

WhereIsIt, an object reminder assistant for blind and severely visually impaired people initiated by Prof. Dr. Oliver Bendel and implemented by Damian Huckele, reached its mid-term presentation on May 19, 2026. The inclusive AI project focuses on a lightweight speech-based system that allows users to store and retrieve object locations without continuous camera use. The work completed so far includes the literature review, problem analysis, requirements definition and the first system design concept. An important milestone was an expert interview with Steve Weidel, blind developer, founder of INKLUTEC and specialist in accessibility and assistive AI technologies. The interview confirmed the practical relevance of the project and highlighted key requirements such as simple voice interaction, portability, reminder functions and optional Bluetooth tags with acoustic tracking. The planned prototype architecture is Python-based and combines speech recognition, language processing, local object-location storage and text-to-speech output. The next project phases include prototype development, testing and the evaluation of Bluetooth integration.

Posted on

A Successful Conference on GenAI and Sexuality

Those who had already arrived on Sunday or Monday were welcomed to Montreal with beautiful sunshine and mild temperatures. The Clock Tower Beach had been prepared, though it was not yet open. On Thursday, April 30, 2026, the SAGA conference took place at the Judith-Jasmin Pavilion Extension in slightly cooler weather. Renowned experts such as David Lafortune and Simon Dubé were among the hosts, alongside rising stars like Valérie A. Lapointe. Through a series of outstanding presentations and panels, the conference explored the theme “Sexuality and Generative AI: Benefits, Risks, and Paths for Action”. This included both generative AI and other AI systems on computers, as well as those embedded in robots and physical systems. The talks, delivered in English and French, were translated live with the help of AI. Both on stage and in the audience were sexologists, psychologists, and philosophers, as well as practitioners from a wide range of fields. There was no hesitation – neither among participants nor in engaging with the topics discussed. This conference was made possible through the financial support of the Social Sciences and Humanities Research Council (SSHRC). The program can be viewed here.

Posted on

Chatbots for Endangered Languages

On April 8, 2026, the article “Chatbots for Dead, Endangered, and Extinct Languages: Possibilities and Limitations of Generative AI for Continuing Education” by Oliver Bendel was published in Wiley Industry News. The focus is on how chatbots based on generative AI can contribute to the preservation and promotion of dead, endangered, and extinct languages in continuing education (as well as in vocational training). Following an introduction to the technical and conceptual foundations, several projects at the FHNW School of Business are presented and discussed from technical, ethical, and didactic perspectives. These dimensions are revisited in the next section and expanded into general and overarching considerations. Finally, possible and necessary steps are outlined that go beyond the purely technological discourse. Additionally, an outlook is provided on future possibilities related to new versions of large language models. This article provides the first comprehensive overview of the projects initiated by Oliver Bendel that are dedicated to dead, endangered, and extinct languages, including @ve (for Latin), @llegra (for Vallader), and kAIxo (for Basque), as well as Cleop@tr@ (Egyptian). It can be accessed via the publisher’s website or downloaded here as a PDF.

Posted on

A Virtual Buddy for Inmates

Inmates in Swiss correctional facilities have only limited access to digital technologies and little contact with the outside world. At the same time, they spend several hours each day in their cells. The project “Chatbots for Inmates” (short title) aims to address this situation: an LLM-based chatbot is intended to serve as a virtual buddy that enables conversations, reduces loneliness, and strengthens social and digital skills. The project was initiated by Tamara Siegmann, founder of the start-up SIEG – Smart Innovation Engineering Group by Siegmann and a student at the FHNW School of Business. She serves as project leader; Prof. Dr. Oliver Bendel is responsible for the project, and Stephan Vonschallen is a project collaborator. The application partner is Lenzburg Prison (JVA Lenzburg), represented by its director Marcel Ruf. The chatbot will run offline on tablets used by the inmates. In a pilot phase, the solution will initially be tested on two or three tablets and later on 30 tablets. The project will investigate whether a digital conversational partner can improve inmates’ well-being and support them in developing skills for life after release. The project began on March 1, 2026, and is supported by an innovation voucher (Innoscheck) from Innosuisse as well as by the Institute for Information Systems at the FHNW School of Business. Back in 2024, Tamara Siegmann and Oliver Bendel carried out a project on Swiss prisons, investigating whether collaborative and social robots could and should be used there. The paper can be downloaded or purchased here (Photo: Boehlich, Wikimedia, CC0 1.0 Universal).

Posted on

AI and Human Creativity

As part of the AAAI Spring Symposia, the symposium “Will AI Light Up Human Creativity or Replace It?: Toward Well-Being AI for co-evolving human and machine intelligence” focuses on how advances in generative AI, large language models, and multi-agent systems are transforming human creativity and decision-making. It addresses the central question of whether AI will amplify human potential or increasingly take its place. The symposium advances the idea of Well-Being AI, emphasizing human-AI collaboration and co-evolution rather than AI as an isolated or autonomous system. While highlighting the potential of AI to support creativity, discovery, and personal development, it also examines risks such as overreliance, reduced diversity of thought, and loss of human autonomy. Chaired by Takashi Kido (see photo) of Teikyo University and Keiki Takadama of The University of Tokyo, the symposium brings together researchers and practitioners from technical, philosophical, and social disciplines to discuss principles and frameworks for AI that augments rather than replaces human creativity. Further information on this symposium and the broader AAAI Spring Symposia organized by the Association for the Advancement of Artificial Intelligence can be found at https://aaai.org/conference/spring-symposia/sss26/ and webpark2506.sakura.ne.jp/aaai/sss26-will-ai-light-up-human-creativity-or-replace-it/.

Posted on

Sexuality and Generative AI

The upcoming “SAGA: Sexuality and Generative AI” symposium, taking place on April 30, 2026 at the Université du Québec à Montréal, explores how generative AI is reshaping intimacy, desire, relationships, and sexual expression. As AI systems increasingly create images, stories, and even interactive romantic companions, they raise new ethical, legal, and social questions that researchers, practitioners, and industry voices will tackle together. The hybrid, bilingual event is free and open to all, offering live translation and captioning. In its first block on social impacts, the symposium features talks such as Brian Willoughby’s exploration of how romantic AI affects young adults’ relationships, Arnaud Anciaux’s analysis of ownership and consent around AI-generated sexual content, Beáta Bőthe’s work on AI-generated pornography and problematic use, and Oliver Bendel’s look at erotic chatbot interactions. With discussions ranging from the regulation of AI-generated pornography to digital intimacy and AI-supported therapeutic interventions, the event promises a rich and essential conversation. Full details and submissions are available at https://event.fourwaves.com/sexualiteia/pages.

Posted on

AI Systems Harm the German Language

Users who translate texts from English or another language into German and are not native speakers of the target language should be cautious when using services such as DeepL and ChatGPT. 1. For both, the default setting is not the standard language, as one might assume, but a special language that is rejected by the majority of the language community and does not follow the official rules. These are determined for all German-speaking countries by the Rechtschreibrat. DeepL and ChatGPT follow their own rules or the inconsistent ideas of activists. The German language generated by DeepL and ChatGPT is often dysfunctional, incorrect, and imprecise. Formal inaccuracies can lead to inaccuracies in content. 2. If AI systems do not know words, they may simply replace them with completely different ones. In one test, DeepL translated “Animal-Computer Interaction” as “Mensch-Computer-Interaktion” (“Human-Computer Interaction”). This made the text factually incorrect. 3. Overall, especially with ChatGPT, English language structures are transferred to German. This results in unnatural-sounding lists, unlinked compounds (“Deep Learning Modelle” or “Deep Learning-Modelle” instead of “Deep-Learning-Modelle”), and unnecessary or incorrect hyphens (“nicht-amtliche Regeln” instead of “nichtamtliche Regeln”).

Posted on

GPT-AfterDark is Coming

According to several media reports on 15 October 2025, ChatGPT is set to get an erotic function. This is likely to include features such as dirty talk – via text and voice – but possibly also instructions for all kinds of positions and tips and tricks for sex toys and sex robots. This follows in the footsteps of other chatbots such as Replika. However, these often have an avatar to make them irresistible. This is not the case with ChatGPT, apart from the small round tiles of the GPTs, the “custom versions” that anyone can easily create. Among these, incidentally, is a SexGPT by Dominick Pandolfo – ‘Provides sexual health information’, so quite harmless. Artificial Life’s virtual girlfriend already existed at the turn of the millennium, also in linguistic and visual form. If OpenAI does not improve this, users will build something themselves, which is already being done today, albeit not necessarily in a sexual sense. Meshy AI and Co. can be used to generate and animate three-dimensional avatars. It will be interesting to see whether the German ChatGPT version uses gender language in its erotic function – as it does in the default setting. Some people may find this arousing, others may not. When asked what this version of ChatGPT could be called, the chatbot itself suggested: ChatGPT Red, GPT-AfterDark, or DeepLure. If that doesn’t turn you on, there’s no helping you.

Posted on

AAAI 2026 Spring Symposium Series

On September 4, 2025, the Association for the Advancement of Artificial Intelligence (AAAI) announced the continuation of the AAAI Spring Symposium Series. The symposium will be held from April 7–9, 2026, at the Hyatt Regency San Francisco Airport in Burlingame, California. The call for proposals for the symposium series is available on its website. According to the organizers, proposals are due October 24, 2025, and early submissions are encouraged. “The Spring Symposium Series is an annual set of meetings run in parallel at a common site. It is designed to bring colleagues together in an intimate forum while at the same time providing a significant gathering point for the AI community. The two and one-half day format of the series allows participants to devote considerably more time to feedback and discussion than typical one-day workshops. It is an ideal venue for bringing together new communities in emerging fields.” (AAAI website). As was the case this year, the Spring Symposium Series will once again not be held on the Stanford University campus. For many years, the History Corner served as the traditional venue for the event. Efforts to secure an alternative university location in the Bay Area have been unsuccessful. AAAI should seriously consider returning to Stanford in 2027. Only then can the Spring Symposium Series regain the atmosphere and significance it once enjoyed.