Perplexity is building an AI phone

Hello everyone! Here’s what you need to know about AI today:

👉 T-Mobile’s parent company and Perplexity are building an AI phone

👉 Opera revealed a new agentic AI web-browsing feature

👉 Gemini now lets you ask questions using videos

and many more!

📧 Did someone forward you this email? Subscribe here for free to get the latest AI news everyday!

Read time: 5 minutes

What’s going on: Deutsche Telekom (DT), the parent company of T-Mobile, revealed its plans to launch an “AI Phone” in collaboration with Perplexity alongside other partners like Picsart. This new smartphone aims to integrate advanced AI deeply into its functionality, promising a seamless user experience driven by AI agents rather than traditional app-based navigation. The device, set to be unveiled in the second half of 2025 and available for purchase in 2026, will initially target the European market.

What does it mean: This partnership builds on a concept first introduced at last year’s MWC, signaling a move into AI-driven hardware. The AI Phone will incorporate additional AI services from Google Cloud and ElevenLabs aiming to offer capabilities like real-time translation, voice synthesis, and image editing, all integrated into a cohesive system. Perplexity’s CEO, Aravind Srinivas, emphasized the shift from reactive AI to proactive assistance, envisioning a device that can book flights, make reservations, and manage communications autonomously.

More details: 

• Specific hardware details such as the operating system or manufacturer remain unknow, but concept renderings suggest a design resembling an Android-based device with a price tag of less than $1,000.

• Previously, Perplexity launched an Android assistant in January that seems like it could be a likely template or inspiration for its new AI Phone.

• Deutsche Telekom's Magenta AI, an app-based version of their AI assistant, will be accessible to existing customers who can install it on their Android or iOS devices, provided they are among DT's 300 million subscribers.

What’s going on: Opera has unveiled a new feature in web browsing with the introduction of its “Browser Operator”. This native AI agent which is integrated directly into the Opera browser, is designed to transform the way users interact with the internet by autonomously performing tasks based on natural language prompts. Unlike traditional browsers that serve as platforms for accessing content, Browser Operator is trying to shift the paradigm, evolving the browser into an active assistant capable of executing practical online activities.

What does it mean: This feature allows users to perform tasks such as shopping, booking tickets, or gathering information automatically. The Browser Operator processes commands locally on the user's device rather than relying on cloud-based systems or virtual machines, as some competing AI products like OpenAI’s Operator do. This client-side approach ensures that sensitive data, including login credentials and browsing history, remains secure and is only shared with the websites explicitly involved in the task.

More details:

• Opera highlights that the AI does not depend on screenshots or video captures to interpret web content; instead, it understands the underlying structure of webpages.

• Users maintain full control throughout the process, with the ability to review each step the AI takes, intervene if necessary, and provide additional input for sensitive actions like entering payment details.

• This feature marks Opera as the first major browser to integrate a web-browsing agent in the 30-years-old history of the existence of web browsers.

• Other companies are also following the same path. The Browser Company, which makes the Arc Browser, has teased a new browser called Dia that will have agentic capabilities, and Perplexity is preparing to launch its own browser, called Comet.

• Interested to see how it works? Watch its demo on YouTube.

💰 Anthropic, the company behind Claude, has raised another $3.5 billion to pursue AI ambitions bringing its total funding to $18.2 billion and valuing the company at $61.5 billion.

🔒 Google Gemini users can now access the AI chatbot directly from their iPhone's lock screen via a new widget, offering features like Gemini Live voice access and quick actions for photos, reminders, and text chats.

💻 Chinese buyers are obtaining Nvidia's Blackwell chips through third-party traders in regions like Malaysia, Taiwan, and Vietnam, who are reselling them to companies in China, raising concerns about the effectiveness of US export restrictions.

💸 TSMC has committed to investing $100 billion to expand its chip manufacturing facilities in the United States.

📽 Google is adding "Screenshare" and video question features to Gemini on Android, allowing users to share their phone screen or ask questions about videos in real-time. These features will roll out to Gemini Advanced users on the Google One AI Premium plan later this month.

👩‍💻 Google has upgraded Colab by integrating its Data Science Agent, powered by Gemini 2.0 AI, to help users clean data, visualize trends, and gain insights from datasets.

👮‍♂️ Singaporean police arrested three men (two Singaporeans and one Chinese citizen) for allegedly smuggling Nvidia chips. Singapore is also investigating whether servers shipped to Malaysia contained restricted Nvidia chips that were potentially diverted elsewhere.

AI + Education

Snorkel AI - Data Annotator - Science, Technology, Engineering, Math

Hippocratic AI - Staff Software Engineer

Kenvue - Senior Data Scientist

Amazon - Machine Learning Engineer, CreativeX

Thank you for staying with us like always! If you are not subscribed, subscribe here for free to get more of these emails in your inbox! Cheers!