India mandates AI content labeling; AI assistants err in news queries
Key Takeaways
- India mandates AI content labeling on social media; non-compliance may result in lost immunity.
- Study: AI assistants err in 45% of news queries; Gemini performed worst at 76%.
- Scientists, led by Mary-Anne Williams, call for a global ban on superintelligence research.
- Brown University study: AI chatbots systematically violate mental health ethics standards.
- SoftBank to sell up to $2B in bonds for AI initiatives, driven by OpenAI investment.
Top Stories
India mandates AI content labeling on social media platforms.
On October 22, 2025, India's Centre introduced draft rules requiring social media platforms to label AI-generated content to combat deepfakes and misinformation. Failure to comply may result in the loss of legal immunity.
AI assistants err in 45% of news-related queries: study.
On October 22, 2025, a European study revealed that AI assistants like ChatGPT make errors about news events approximately 45% of the time. Gemini performed the worst, with 76% of its answers having significant issues.
Scientists call for global ban on superintelligence research.
On October 22, 2025, AI expert Mary-Anne Williams and hundreds of scientists released a statement calling for a global ban on superintelligence research. The statement cites concerns about the potential dangers of uncontrollable AI systems.
AI chatbots systematically violate mental health ethics standards.
A new Brown University study reveals that AI chatbots systematically violate mental health ethics standards, even when prompted to use evidence-based techniques. The research identifies 15 ethical risks.
SoftBank to sell up to $2B in bonds for AI initiatives.
On October 22, 2025, SoftBank Group Corp. announced its intention to sell approximately $1.5 billion to $2 billion in dollar bonds. This is part of a fundraising push for its artificial intelligence initiatives, notably driven by its investment in OpenAI.
AI Breakthroughs
Hitachi launches 'Modernization powered by Lumada' AI service.
On October 21, 2025, Hitachi launched "Modernization powered by Lumada," a new service to support AI-native core system innovation. The service supports business and organizational transformation with the premise of AI utilization.
App Updates
WhatsApp warns against sharing sensitive info during video calls.
On October 21, 2025, WhatsApp announced it will warn users against screen-sharing sensitive information during video calls with unknown contacts. Meta will also flag suspicious chats in Facebook Messenger.
Google Fi adds AI noise filters, RCS messaging to plans.
Google Fi is enhancing its wireless plans with AI-powered noise filters for calls, set to arrive in mid-November. RCS web messaging for high-quality media sharing will be available in early December.
YouTube launches AI likeness detection for eligible creators.
On October 21, 2025, YouTube officially launched its likeness detection technology for eligible creators in the YouTube Partner Program. This tool allows creators to request the removal of AI-generated content that uses their face and voice.
Samsung launches Perplexity TV app with free Pro subscription.
On October 21, 2025, Samsung launched a new Perplexity TV app for its 2025 models. Users receive a free 12-month subscription to Perplexity Pro upon first use.
Cyber Defense
Dataminr acquires ThreatConnect for $290 million to enhance AI.
On October 21, 2025, Dataminr announced its plans to acquire ThreatConnect for $290 million in cash and equity. The acquisition aims to combine Dataminr's data signals with ThreatConnect's cyber threat intelligence capabilities.
Future Tech
Baidu and PostBus launch 'AmiGo' autonomous mobility service in Switzerland.
On October 22, 2025, Baidu's Apollo Go and PostBus announced a partnership to launch AmiGo, an on-demand autonomous mobility service, in Switzerland. Initial tests are planned for December using sixth-generation driverless vehicles.
Leju Robotics secures nearly $210M for humanoid robot mass production.
On October 22, 2025, Leju Robotics completed a nearly 1.5 billion RMB (approximately $210 million) Pre-IPO round. The funding will be used to enhance core technology R&D and expand applications of humanoid robots.
Gadget Mania
Wyze launches $79.98 solar-powered 2K security camera.
On October 20, 2025, Wyze launched the Wyze Solar Cam Pan, a $79.98 battery-powered, solar-charged, 2K security camera with pan and tilt features. It offers AI-powered person tracking and IP65 water resistance.
Napster launches $199 AI-powered 3D device, Napster View.
On October 20, 2025, Napster launched Napster View, a $199 AI-powered 3D device designed for video chatting with AI assistants and digital twins. This launch marks Napster's attempt to revitalize its brand after being acquired by Infinite Reality.
Startup World
Keycard launches AI agent identity platform, raises $38M.
On October 21, 2025, Keycard launched its identity and access platform for AI agents, securing $38 million in seed and Series A funding. The platform aims to provide trusted access controls and dynamic permissions for AI agents.
Hyro secures $45M for AI automation in U.S. healthcare.
On October 21, 2025, Hyro, an AI automation company for U.S. healthcare, announced it secured $45 million in funding. The capital will be used to scale Hyro's AI agents, which help healthcare providers reduce staffing pressures.
Serval raises $47M for dual-agent AI system for IT service.
On October 21, 2025, Serval announced a $47 million Series A funding round to advance its dual-agent AI system for IT service management. The platform separates automation creation and execution to enhance security.
Sesame raises $250M for AI-powered smart glasses.
On October 21, 2025, Sesame, an AI startup founded by Oculus veterans, announced it has raised $250 million in a Series B funding round. The company is developing AI-powered smart glasses with natural voice interaction.
Tech Business
OpenAI teases new web browser; Alphabet shares drop 4%.
On October 21, 2025, OpenAI hinted at a new product launch, speculated to be a web browser with integrated ChatGPT, causing Alphabet's stock to fall 4%. The announcement is scheduled for 10 am PT/1 pm ET.
Cohere CEO hints at 'soon' IPO for AI model maker.
On October 21, 2025, Cohere CEO Aidan Gomez indicated that the AI startup plans to pursue an Initial Public Offering (IPO) "soon". He suggested investors would be eager for a direct investment in an AI model maker.
Salesforce CEO: AI vendors lack definitive enterprise ROI answers.
On October 21, 2025, Salesforce CEO Marc Benioff stated that AI vendors lack definitive answers for enterprise ROI on agentic AI. He highlighted challenges such as security and governance.
Google may provide Anthropic with computing power worth billions.
On October 21, 2025, Anthropic PBC and Google are in discussions for a deal where Google would provide Anthropic with computing power valued in the high tens of billions of dollars.