Tech Roundup: E.U. AI Act, Samsung Galaxy Ring & More

[A recurring feature on the latest in Science & Technology.]

• The European Union publishes the final text of the AI Act, a landmark risk-based regulation for applications of artificial intelligence; to come into effect on August 1, 2024, with companies having time till 2026 to comply with all its provisions. (That said, six months from now, the bloc will start implementing bans on prohibited applications for AI, such as the use of social credit ranking systems, the collection and compilation of facial recognition information for databases, as well the use of real time emotion recognition systems in schools and workplaces.)
• The European Commission finds that X's paid blue checkmark deceives users and breaches the Digital Services Act areas linked to dark patterns, advertising transparency and data access for researchers; says it "negatively affects users' ability to make free and informed decisions about the authenticity of the accounts and the content they interact with."
• U.S. lawmakers introduce a new bill called the Content Origin Protection and Integrity from Edited and Deepfaked Media Act (COPIED Act) that seeks to protect artists, songwriters and journalists from having their content used to train AI models or generate AI content without their consent.
• Nothing unveils the US$ 199 CMF Phone 1, with swappable parts, a 6.67" AMOLED display, a MediaTek SoC and Nothing OS, alongside the US$ 69 CMF Watch Pro 2 and US$ 59 CMF Buds Pro 2.
• Amazon announces a new US$ 79 Echo Spot alarm clock with a revamped design and improved audio quality; launches App Studio in AWS, a generative AI-powered service letting enterprise users create internal apps using text prompts.
• Open-source blogging platform and Substack rival makes its own newsletter service available on the fediverse via ActivityPub integration, joining other services like Mastodon, Pixelfed, PeerTube, Flipboard and Instagram Threads.
• The U.S. Federal Trade Commission (FTC) bans anonymous messaging app NGL from offering its services to children under the age of 18 and halt deceptive claims that its AI content moderation program filtered out cyberbullying and other harmful messages; says the company "sent fake messages that appeared to come from real people and tricked users into signing up for their paid subscription by falsely promising that doing so would reveal the identity of the senders of messages." (The development comes as the consumer protection agency said that websites and mobile apps are increasingly using dark patterns that aim to "manipulate consumers into buying products or services or giving up their privacy.")
• Exercise equipment and media company Peloton faces a class-action lawsuit over accusations that it allowed a third-party marketing firm Drift to process chat data between Peloton users and company representatives without their permission; says the web chats are being used to improve Drift's own AI systems, once again underscoring how companies finding varied ways to gather training data.
• Samsung unveils Galaxy Ring that aims to help users "manage your health by continuously measuring and analysing various vital signs and activity patterns in detail" for US$ 400; also debuts Galaxy Z Fold6, Galaxy Z Flip6, Galaxy Buds 3, Galaxy Buds 3 Pro, Galaxy Watch7 and Galaxy Watch Ultra (both the smartwatch models will ship with Google Wear OS 5) at its annual Unpacked event.
• Amazon expands its AI shopping assistant Rufus in the Amazon mobile app for all U.S customers, after launching in beta for select customers in February 2024; gets hit with a new lawsuit in the U.K. for allegedly abusing its dominant position to favour its own retail offerings and logisics services.
• Spotify adds the ability for users to leave comments on podcasts, expanding on polls and Q&As (comments will private by default and creators can opt out), as it attempts to go beyond streaming to become a social network centered around audio; comes as the company faces the heat over a controversial decision to reclassify its Premium tiers as bundles by combining music and audiobooks, which has been described as "a scheme to increase profits by deceiving consumers and cheating the music royalty system."
• Meta's WhatsApp allows businesses to send authentication codes to users in India and brings voice message transcription (in English, Spanish, Portuguese, Russian and Hindi) to Android a year after its debut on iOS, and makes it easier for users to exit suspicious group chats with a new context card; tests on-device language translations in chat messages powered by Google Translate and letting Community admins hide their groups from the public eye.
• Apple launches the Apple Vision Pro in Australia, Canada, France, Germany and the U.K. after launching it in China, Hong Kong, Japan, and Singapore; settles with the European Commission over Apple Pay and avoids potential antitrust fines by agreeing to open NFC access on iOS to third-party wallet developers for free for a decade.
• Google begins rolling out speedometers in Google Maps for iOS and CarPlay globally, more than five years after adding the features to the Android app; partners with Apple as part of their data portability initiative to let Google Photos users transfer their collections directly to iCloud Photos.
• OpenAI joins hands with Thrive Global to build a hyper-personalised AI health coach that aims to offer "real-time nudges and recommendations unique to you that allows you to take action on your daily behaviors to improve your health."
• Google updates Gemini for Android to work better with foldables, expands Circle to Search by supporting solutions for symbolic math equations and scanning of barcodes and QR codes; experiments with an Instagram-like "Add yours" prompt stickers in YouTube Shorts and a new, invite-only "My Week" feature in Google Photos that allows users to "forget about followers" and share a "weekly photo journal for you and a few special people you choose."
• Swiss privacy company Proton launches Secure Links in Proton Pass to safely share passwords.
• Online marketplace Etsy revises its seller policy to introduce new labels that clarify whether a seller made, designed, sourced or handpicked an item, as it attempts to tackle an influx of AI-generated products on its platform.
• Video hosting and sharing service Vimeo joins TikTok, YouTube and Meta in implementing a way for creators to label AI-generated content. (The company previously said it "will not allow generative AI models to be trained using videos hosted on our platform without your explicit consent, even if you use our free offerings.")
• Google discontinues Play Service updates for Android Lollipop (version 5.0), ending a decade-long support for the operating system version.
• AI company Anthropic now lets developers use Claude 3.5 Sonnet to generate, test and evaluate their prompts; OpenAI creates an internal scale comprising five levels to track its progress toward artificial general intelligence: Chatbots, Reasoners, Agents, Innovators and Organisations, and says it's nearly at level two. (OpenAI has also come under scrutiny from its employees who allege the company "illegally prohibited its employees from warning regulators about the grave risks its technology may pose to humanity" by making them sign restrictive agreements.)
• Microsoft announces plans to hike its Xbox Game Pass Ultimate pricing in September 2024 to US$ 19.99 a month, alongside launching a new standard subscription that doesn't include day-one access to first-party Xbox games.
• Global PC shipments rise 3% YoY to 64.9 million units in Q2 2024; Apple's shipments incease 20.8% YoY, the biggest jump among global PC makers.
• Apple faces a new regulatory setback after the Indian antitrust body CCI finds that Apple exploited its dominant App Store position on iOS and engaged "in abusive conduct and practices"; company denies any wrongdoing and claims it's a "small player in India."
• The U.S. Justice Department sues Adobe for allegedly harming "consumers by enrolling them in its default, most lucrative subscription plan without clearly disclosing" plan terms.