AI News of the Week: Global Highlights from April 20–27, 2025

Welcome to our weekly AI news of the week roundup! The period of April 20–27 was packed with global AI news – from government initiatives and legal battles to cutting-edge tech launches. We’ve collected the most notable international stories and presented them in a playful, easy-to-scan format. Let’s dive in!

AI News in Education and Media: White House Initiative & Italian News Experiment

• US Launches AI Education Push: President Trump signed an executive order establishing a national AI education initiative and a White House Task Force on AI Education. The plan will channel resources into AI training in schools, teacher development programs, and even a new AI-focused Presidential Challenge to spur innovation.
  Source: The Washington Post
• Italian Newspaper Embraces AI Writers: Italian daily Il Foglio found success with AI-driven journalism. They published a daily four-page section entirely written by AI for a month and saw a boost in sales, so now it’s becoming a weekly feature. The editor noted the AI writes with plenty of irony and can summarize books well – though it still lacks human-like critical thinking.
  Source: Reuters

YouTube Tests AI Features in Search and Music

• AI Video Summaries in Search: YouTube is experimenting with an “AI Overviews” carousel that surfaces short video clips relevant to a user’s search query. Only a small number of US-based Premium users see this AI-generated summary feature (for select English queries like product reviews or travel), but it mirrors Google’s own AI summaries in Search which have been controversial for diverting traffic.
  Source: MacRumors
• AI-Generated Music for Creators: YouTube also introduced a new “Music Assistant” tool that generates copyright-free background music from text prompts. Available to some users in the Creator Music beta, it can whip up multiple instrumental tracks for download and editing – helping creators easily add legal, custom music to their videos.
  Source: TechCrunch

China’s AI Advances: Baidu’s Chatbot Upgrades & Kuaishou’s Video Generator

• Baidu Unveils Cheaper Ernie Models: Chinese tech giant Baidu rolled out two improved versions of its Ernie AI chatbot to stay ahead in the AI race. The new Ernie 4.5 Turbo model is about 80% cheaper to run than its predecessor, and a reasoning-focused Ernie X1 Turbo is 50% cheaper. These cost cuts come as Baidu faces rising competition from local startups and global rivals like OpenAI.
  Source: Reuters
• Kuaishou’s Powerful Video AI: Kuaishou – the company behind one of China’s top short-video apps – unveiled Kling 2.0, calling it the world’s most powerful AI video generator. The upgraded system boasts more realistic visuals and smoother motion, and it has already churned out over 168 million AI-generated videos (plus 344 million images) to date! Kuaishou’s claim to the AI video crown shows China’s firms pushing the envelope in generative media tech.
  Source: YourStory

AI’s Impact: Productivity Gains and Economic Growth

• Workers Save Time with AI: A Google pilot program in the UK found that using AI for tedious admin tasks could save workers an average of 122 hours per year (that’s roughly three extra work weeks of time saved). After a bit of training and giving employees “permission to prompt,” AI adoption in the test group surged – weekly AI use jumped from 17% to 56% among older workers once they felt more confident using the tools.
  Source: Reuters
• IMF Sees AI Lifting the Economy: The International Monetary Fund estimates that AI could boost global GDP by about 0.5% annually from 2025 to 2030. The added productivity outweighs the increased emissions from power-hungry data centers, the IMF says. However, the gains won’t be evenly distributed – the IMF urges sustainable AI deployment so benefits are shared broadly (and climate impacts are mitigated).
  Source: IMF

OpenAI’s Big Moves: Japan Partnership, In-Chat Shopping, and Browser Ambitions

• Enterprise AI in Japan: OpenAI partnered with Japan’s NTT Data to bring ChatGPT Enterprise to Japanese businesses and co-develop advanced AI solutions. NTT will launch an OpenAI Center of Excellence and help 100 companies jumpstart AI projects, with the partnership aiming for nearly $700 million in related revenue by 2027.
  Source: Capacity Media
• Shopping via ChatGPT: In e-commerce news, code sleuths discovered hints of a Shopify integration inside ChatGPT, suggesting users may soon browse, review, and purchase products without leaving the chat interface. This native in-chat shopping feature appears close to launch and would mirror similar moves by Microsoft and others to enable “conversational commerce”. Source: AutoDS
• Eyeing a Web Browser: OpenAI is even contemplating the web browsing market. The company’s product chief said they’d consider buying Google Chrome if antitrust rulings forced Google to sell it. OpenAI has already hired some former Chrome engineers and toyed with building its own browser – hinting at a future where ChatGPT could be deeply integrated into how we surf the web. Source: Business Insider

AI at Work: New Microsoft Copilot Agents and HubSpot’s AI Boost

• Microsoft 365 Gets AI Agents: Microsoft unveiled two new AI helpers for Microsoft 365 users, dubbed “Researcher” and “Analyst,” as part of its Copilot suite. These agents can handle heavy-duty workplace tasks – think automatic meeting summaries, drafting reports, or analyzing CRM sales data – and will be accessible through an upcoming “Agent Store” in Microsoft 365.
  Source: Microsoft Blog
• HubSpot Acquires AI Knowledge Startup: CRM platform HubSpot is also doubling down on AI, announcing the acquisition of Dashworks, a startup specializing in deep search and knowledge management. The goal is to power up HubSpot’s own AI assistant (called Breeze) so it can retrieve unstructured info across a company’s data via natural language. In practice, this means marketing and sales teams could ask Breeze a question and get an instant answer pulled from internal documents, wikis, emails, and more.
  Source: HubSpot News

Creative Tools Get an AI Boost

• Design Apps Integrate ChatGPT’s Image AI: Adobe and Figma are integrating OpenAI’s newest image-generation model into their creative tools. OpenAI’s powerful “GPT-Image” model (now available via API) lets users generate and edit images from text prompts right inside those apps, and other platforms like Canva and GoDaddy are exploring similar features.
  Source: The Verge
• Adobe Upgrades Firefly (and Plays Nice with Others): Adobe rolled out its own Firefly Image Model 4 (plus a high-octane Model 4 Ultra) to deliver faster, higher-quality AI art generation. In a twist, Adobe’s web app now lets creators use third-party AI models – from OpenAI, Google, and others – for image and video generation alongside Adobe’s models. Adobe assures that its in-house models are trained on licensed content and thus “commercially safe” for businesses, addressing concerns about copyright.
  Source: Adobe News

AI & Content Rights: Lawsuit and Creator Protections

• OpenAI Sued Over Copyrights: The AI vs. content creator battle heated up as Ziff Davis – owner of IGN, PCMag, Mashable and more – sued OpenAI for allegedly misusing its content. The lawsuit claims OpenAI scraped articles from Ziff Davis’s sites (ignoring robots.txt rules) and even stripped out copyright notices when feeding the data into its models. Ziff Davis joins other publishers (like The New York Times) in legal action over how ChatGPT is trained.
  Source: Reuters
• Creators Guild “AI Rider” Contract: On the flip side, content creators got a new tool to protect themselves. The Creators Guild of America (CGA) introduced a binding contract add-on