Top AI Developments of the Week From European Union's first draft of a Code of Practice for general-purpose AI, to Google's Gemini AI for iOS

Google Gemini AI Launches for iOS

Google has debuted a standalone Gemini AI app for iPhones, offering a conversational assistant akin to Siri. Named Gemini Live, the app integrates seamlessly with services like YouTube, Google Maps, and Gmail for real-time information retrieval and customizable chatbot interactions. Available in twelve languages—including Hindi, English, French, German, Spanish, Portuguese, Arabic, Italian, Japanese, Indonesian, Vietnamese and Turkish. It enhances the iOS user experience with a dedicated interface, 10 voice options, and cross-app integration.

"iPhone users can now experience Gemini in a whole new way with our dedicated mobile app," Google explained in a blog post. "In addition to accessing Gemini through the Google app or a web browser, iPhone users can enjoy a more streamlined Gemini experience, with features designed to improve learning, creativity, and productivity."

EU Releases Draft Code for General-Purpose AI

The European Union has unveiled its first draft Code of Practice for General-Purpose AI models under the AI Act, set to take effect in August 2024. This draft targets companies such as OpenAI, Google, Meta, Anthropic, and Mistral, providing a framework for risk management, penalties, and compliance. Fines for violations could reach Euro 35 million (USD 37.19 million) or seven per cent of global annual revenue.

"This first draft of the Code is the result of a preliminary review of existing best practices by the four specialised Working Groups, stakeholder consultation input from nearly 430 submissions, responses from the provider workshop, international approaches (including the G7 Code of Conduct, the Frontier AI Safety Commitments, the Bletchley Declaration, and outputs from relevant government and standard-setting bodies), and, most importantly, the AI Act itself," the drafters noted.

"We emphasise that this is only a first draft and consequently the suggestions in the draft Code are provisional and subject to change. Therefore, we invite your constructive input as we further develop and update the contents of the Code and work towards a more granular final form for May 1, 2025," they further added.

Stakeholders are invited to provide feedback via the Futurium platform until November 28, 2024, ahead of the final draft's release on May 1, 2025.

Google's AI-Powered Scam Detection Rolls Out for Pixel Users

Google has launched an AI-driven Scam Detection feature in beta for Pixel 6 and newer models in the U.S., with plans to extend it to Android devices soon. This feature scans incoming calls for scam indicators, such as urgent financial demands, and provides warnings via screen alerts, vibrations, and audio notifications.

Addressing privacy concerns, Google stated, "No conversation audio or transcription is stored on the device, sent to Google servers, or retrievable after the call." Users can enable or disable the feature in the Phone app settings and even turn it off for specific calls.

Apple's AI-Driven Smart Home Display in Development

Apple is reportedly working on a wall-mounted smart display with a 6-inch vertical screen—designed to resemble two stacked iPhones. Code-named J490, the device will integrate AI capabilities for smart home control and video conferencing. Expected to launch in March 2025, the display may showcase the forthcoming Apple Intelligence platform, with pricing for the high-end version estimated at USD 1,000. It aims to compete with Google's Nest Hub and Amazon's Echo Show.

Particle's AI-Driven News Summaries Gain Attention

News platform Particle is leveraging AI to summarize articles while offering balanced perspectives across the political spectrum. Co-founder Sara Beykpour, a former Twitter employee, emphasized creating an ecosystem that benefits both readers and publishers by simplifying news consumption.

"Sometimes it feels like headlines are all we have time for. We also want to understand more, but faster," Beykpour explained in a Threads post. The app allows users to choose between quick summaries or deeper insights, enabling them to track how a story evolves over time.

"We want to help readers cut through the noise while offering a sustainable, win-win situation for both audiences and content creators," she added.