AI Ethics News — 2026-10-29 (Thu)

Overview

The questions that surfaced were consent and surveillance (complaints over alleged worker surveillance at Scale AI, a lawsuit over an alleged facial recognition misidentification) and creator consent (the Society of Authors' withdrawal from a charter, objecting to Meta's AI training). Accountability for unreviewed AI output and AI-generated fakes ahead of elections also drew attention. (Reuters, ABC News, Brittle Paper, 404 Media, 조선일보 / Chosun Ilbo, Inside Politic) 【Bias, discrimination and fairness】 - Her Campus published a piece on gender bias in AI. (Her Campus) - The Atlanta Journal-Constitution argues that civil rights era surveillance threatened dissent and that AI is a bigger threat now. (The Atlanta Journal-Constitution) - Experts warn that AI bias and misinformation could cloud choices in the upcoming election. (KBAK) - Scherer looks at AI's impact on fairness, efficiency and quality. (Global Arbitration Review) - The headlines do not give the measurement methods or the scale of any disadvantage.

Privacy and surveillance

- A privacy group filed complaints against Scale AI over alleged worker surveillance. The headline does not say what data was collected or whether workers consented. (Reuters) - A grandmother who was falsely accused of bank theft after an alleged AI facial recognition match is suing for $10M. (ABC News) - Focus groups reveal deep caution among experts on AI chatbot therapists. (Bioengineer.org) 【Transparency and accountability】 - OpenAI released AI math proofs that had not been reviewed, sparking a debate over accountability. The headline does not say what else was or was not disclosed. (Chosun Ilbo) 【Creators, copyright and consent】 - The Society of Authors withdrew from the Global Ethical AI Charter less than two weeks after its launch. The reason is not stated in the headline. (Brittle Paper) - Users in the UK or EU can object to Meta's AI training. 404 Media published a guide on how. (404 Media)

Disinformation and deepfakes

- Confirmed by a fact check: an image claiming to show fertility signs is AI-generated, and its labels are misspelled. (Yahoo) - Concerns have been raised about AI-generated fake news ahead of local government elections. (Inside Politic) - Countermeasure: iOS 27.2 adds Google's hidden SynthID watermark to AI images on iPhone. (Beebom Gadgets) - Claim, not confirmed: a Deputy Commissioner says the voice on a recording with an alleged rape victim is AI-generated and not his. (News Source Guyana) - EFF published a piece on social media when elections are close, titled "Social Media When Elections Are Close: Doing the Impossible Better". (EFF) 【Protecting children and vulnerable people】 No notable developments. 【Ethical principles and guidelines】 - No institution's decision appears in the headlines. TechTarget published guidance on how executives can build a responsible AI framework, which is a proposal rather than a decision. (TechTarget)

AI ethics in Japan

- A headline reports that Suno, the music generation AI, and a US copyright management organization are involved in a request over DSP playback data. The headline does not make clear who is asking whom. (MEDIAMIXI) - Five companies, including Google, responded to Japan's government on measures to ensure transparency of generative AI data. The headline does not give the contents of the responses. (Asahi Shimbun via Yahoo! News)

Statements by key figures

No notable developments. 【Integrated analysis: how the ethical questions connect】 - Missing consent appears in several places: alleged surveillance of Scale AI workers, the need for UK and EU users to object to Meta's AI training, and the Society of Authors' withdrawal from the charter. In each case, the people affected appear to have had little say in how data or work is used. (Reuters, 404 Media, Brittle Paper) - Missing explanation appears in the unreviewed OpenAI proofs and the Japanese government's request for generative AI data transparency. The facial recognition lawsuit suggests that a lack of verifiable grounds for an AI match can lead to harm to an individual. (Chosun Ilbo, Asahi Shimbun via Yahoo! News, ABC News) - Elections link the bias, fake news and deepfake items. Provenance tools such as SynthID and fact checks address the problem after content has been made. The Guyana case shows that claims of AI generation can also be used to dispute a recording. This is inference. (KBAK, Inside Politic, Beebom Gadgets, Yahoo, News Source Guyana) - Companies may face pressure to make opt-out routes, worker data practices and labeling clearer. Users and organizations may need to verify content and AI outputs before relying on them. This is inference.

What to watch

- Whether Scale AI or regulators respond to the privacy group's complaints, and what they reveal about worker data practices. (Reuters) - Whether other creator groups follow the Society of Authors in leaving the Global Ethical AI Charter, and the reasons given. (Brittle Paper) - Whether provenance tools such as SynthID and fact checks affect how AI-generated content is handled in election periods. (Beebom Gadgets, Inside Politic)