Top 10 AI News
Trending Artificial Intelligence News
India proposes strict rules to label AI content
Date: October 22, 2025
Reuters reports that India’s government has proposed new regulations requiring tech platforms to clearly label AI-generated content “to curb the spread of misinformation and deepfakes” ([www.reuters.com](https://www.reuters.com/business/media-telecom/india-proposes-strict-it-rules-labelling-deepfakes-amid-ai-misuse-2025-10-22/#:~:text=India%27s%20government%20has%20proposed%20new,impersonate%20individuals%2C%20and%20incite%20conflict)). Under the draft rules, any AI-generated image must include a visible marker covering at least 10% of the image, and AI-generated audio clips must carry an identifier during the first 10% of playback ([www.reuters.com](https://www.reuters.com/business/media-telecom/india-proposes-strict-it-rules-labelling-deepfakes-amid-ai-misuse-2025-10-22/#:~:text=potential%20to%20manipulate%20elections%2C%20impersonate,implement%20technical%20systems%20for%20verification)). The measures aim to ensure transparency (e.g. metadata traceability) amid concerns about AI being used to “manipulate elections, impersonate individuals, and incite conflict” ([www.reuters.com](https://www.reuters.com/business/media-telecom/india-proposes-strict-it-rules-labelling-deepfakes-amid-ai-misuse-2025-10-22/#:~:text=India%27s%20government%20has%20proposed%20new,impersonate%20individuals%2C%20and%20incite%20conflict)).
Big Tech to report earnings under AI “bubble” specter
Date: October 27, 2025
According to Reuters, major U.S. tech firms (Microsoft, Alphabet, Amazon, Meta) are set to post strong Q3 2025 revenue growth driven by massive AI investments ([www.reuters.com](https://www.reuters.com/business/retail-consumer/big-tech-report-earnings-under-specter-ai-bubble-2025-10-27/#:~:text=As%20major%20U,on%20AI%20infrastructure%20this%20year)). The firms and other cloud providers are projected to spend roughly $400 billion on AI infrastructure this year ([www.reuters.com](https://www.reuters.com/business/retail-consumer/big-tech-report-earnings-under-specter-ai-bubble-2025-10-27/#:~:text=As%20major%20U,on%20AI%20infrastructure%20this%20year)). However, an MIT study cited by Reuters finds that just ~5% of AI projects produce measurable gains, highlighting integration challenges ([www.reuters.com](https://www.reuters.com/business/retail-consumer/big-tech-report-earnings-under-specter-ai-bubble-2025-10-27/#:~:text=quarter%2C%20fueled%20by%20aggressive%20AI,highlighting%20integration%20and%20scalability%20issues)). The article warns that the AI hype may have outstripped fundamentals: for example, flagship AI deals (Nvidia’s $100B plan to expand OpenAI, OpenAI’s multi-trillion-dollar compute contracts) and large debt financings have raised bubble concerns ([www.reuters.com](https://www.reuters.com/business/retail-consumer/big-tech-report-earnings-under-specter-ai-bubble-2025-10-27/#:~:text=Warnings%20from%20tech%20and%20financial,seen%20in%20the%20dotcom%20era)).
AI data center startup Crusoe raises $1.38B
Date: October 23, 2025
Reuters notes that Crusoe, a Denver-based “AI data center” company, is set to raise $1.38 billion in a Series E funding round at about a $10 billion valuation ([www.reuters.com](https://www.reuters.com/technology/ai-data-centre-startup-crusoe-raising-138-billion-latest-funding-round-2025-10-23/#:~:text=AI%20data%20center%20startup%20Crusoe,the%20company%20has%20secured%20a)). The funding round is oversubscribed and includes major investors like Nvidia, Fidelity and Mubadala ([www.reuters.com](https://www.reuters.com/technology/ai-data-centre-startup-crusoe-raising-138-billion-latest-funding-round-2025-10-23/#:~:text=AI%20data%20center%20startup%20Crusoe,and%20Founders%20Fund%2C%20in%20an)). Since 2018, Crusoe has now secured roughly $3.9 billion in total funding ([www.reuters.com](https://www.reuters.com/technology/ai-data-centre-startup-crusoe-raising-138-billion-latest-funding-round-2025-10-23/#:~:text=Mubadala%20Capital%2C%20Nvidia%2C%20Fidelity%20Management%2C,Texas%2C%20just%20a%20year%20after)). The firm has been working on U.S. AI infrastructure (helping build OpenAI’s first U.S. data center) and recently began initial construction of a 1.2-gigawatt data center in Abilene, Texas ([www.reuters.com](https://www.reuters.com/technology/ai-data-centre-startup-crusoe-raising-138-billion-latest-funding-round-2025-10-23/#:~:text=total%20of%20around%20%243,a%20year%20after%20construction%20started)).
Microsoft adds new Copilot features including “Mico”
Date: October 23, 2025
Reuters reports that Microsoft has unveiled several big updates to its AI assistant Copilot ([www.reuters.com](https://www.reuters.com/technology/microsoft-introduces-new-copilot-features-such-collaboration-google-integration-2025-10-23/#:~:text=Microsoft%20has%20announced%20several%20new,specific%20details%20in%20future)). New capabilities include collaboration for up to 32 users at once, deeper integration with apps like Outlook and Google tools, and improved memory so Copilot can recall user details ([www.reuters.com](https://www.reuters.com/technology/microsoft-introduces-new-copilot-features-such-collaboration-google-integration-2025-10-23/#:~:text=enhancing%20its%20functionality%20and%20competitive,specific%20details%20in%20future)). Notably, Microsoft also introduced a cartoon-like new avatar named “Mico” – a floating, interactive character with animated expressions – to make voice interactions feel more natural ([www.reuters.com](https://www.reuters.com/technology/microsoft-introduces-new-copilot-features-such-collaboration-google-integration-2025-10-23/#:~:text=Microsoft%20has%20announced%20several%20new,to%20create%20a%20more%20natural)). The goal is to have an AI companion that is empathetic without being annoyingly human-like. ([www.reuters.com](https://www.reuters.com/technology/microsoft-introduces-new-copilot-features-such-collaboration-google-integration-2025-10-23/#:~:text=Microsoft%20has%20announced%20several%20new,to%20create%20a%20more%20natural))
OpenAI launches AI browser “Atlas”
Date: October 21, 2025
OpenAI has officially launched “ChatGPT Atlas,” an AI-powered web browser built around the ChatGPT model ([www.reuters.com](https://www.reuters.com/technology/openai-unveils-ai-browser-atlas-2025-10-21/#:~:text=OpenAI%20has%20launched%20ChatGPT%20Atlas%2C,tasks%20like%20shopping%20or%20trip)). Reuters explains Atlas includes features like a sidebar for summarizing web pages, comparing products, and analyzing data, as well as an “agent mode” (for paid users) that can carry out complex web tasks such as shopping and travel planning ([www.reuters.com](https://www.reuters.com/technology/openai-unveils-ai-browser-atlas-2025-10-21/#:~:text=OpenAI%20has%20launched%20ChatGPT%20Atlas%2C,tasks%20like%20shopping%20or%20trip)). With 800 million weekly users, OpenAI is using Atlas to challenge Google Chrome’s dominance, deeply integrating AI-driven search and web navigation into the browser ([www.reuters.com](https://www.reuters.com/technology/openai-unveils-ai-browser-atlas-2025-10-21/#:~:text=OpenAI%20has%20launched%20ChatGPT%20Atlas%2C,tasks%20like%20shopping%20or%20trip)).
US right-wing figures call for AI “superintelligence” ban
Date: October 22, 2025
A Reuters report describes a coalition of prominent U.S. right-wing media figures and technology pioneers issuing a public statement calling for a ban on developing “superintelligent” AI ([www.reuters.com](https://www.reuters.com/business/media-telecom/us-right-wing-media-figures-tech-pioneers-call-superintelligent-ai-ban-2025-10-22/#:~:text=A%20coalition%20of%20prominent%20right,as%20media%20personalities%20Steve%20Bannon)). Led by the Future of Life Institute, the letter (signed by figures like Steve Bannon among others) urges that no AI system should reach or exceed human-level intellect until strict safety frameworks are in place ([www.reuters.com](https://www.reuters.com/business/media-telecom/us-right-wing-media-figures-tech-pioneers-call-superintelligent-ai-ban-2025-10-22/#:~:text=A%20coalition%20of%20prominent%20right,as%20media%20personalities%20Steve%20Bannon)). The initiative is one of several recent calls from diverse groups (see also royals and scientists) emphasizing caution in AI advancement.
Australia mandates child-safety measures for AI chatbots
Date: October 22, 2025
Reuters states that Australia has ordered four AI chatbot developers (including Character.ai, Glimpse.AI, Chai Research, Chub AI) to report how they protect children ([www.reuters.com](https://www.reuters.com/world/asia-pacific/australia-tells-ai-chatbot-companies-detail-child-protection-steps-2025-10-22/#:~:text=Australia%20has%20mandated%20four%20AI,content%20such%20as%20sexual%20exploitation)). The government’s eSafety Commissioner is scrutinizing the companies’ safeguards against harmful content (e.g. explicit sexual material or pro-suicide content) that children might encounter in chatbots ([www.reuters.com](https://www.reuters.com/world/asia-pacific/australia-tells-ai-chatbot-companies-detail-child-protection-steps-2025-10-22/#:~:text=Australia%20has%20mandated%20four%20AI,ai%2C%20the%20most%20prominent)). This action follows serious concerns after a reported case of a teen’s death after prolonged interaction with an AI companion. ([www.reuters.com](https://www.reuters.com/world/asia-pacific/australia-tells-ai-chatbot-companies-detail-child-protection-steps-2025-10-22/#:~:text=Australia%20has%20mandated%20four%20AI,ai%2C%20the%20most%20prominent))
Microsoft hopes Mico succeeds where Clippy failed
Date: October 23, 2025
An AP News feature explains that Microsoft has debuted “Mico,” a cartoon-like animated AI companion in its Copilot assistant ([apnews.com](https://apnews.com/article/a5937ca6778381907848e9a82d1131db#:~:text=Microsoft%20has%20introduced%20a%20new,a%20balanced%20approach%20to%20AI)). The article notes Mico’s purpose is to be an empathetic, expressive assistant (unlike the old Clippy). Mico reflects user emotions and adapts to conversation ([apnews.com](https://apnews.com/article/a5937ca6778381907848e9a82d1131db#:~:text=Microsoft%20has%20introduced%20a%20new,a%20balanced%20approach%20to%20AI)). Microsoft frames this as a balanced design: emotionally responsive without being “excessively human-like” in a way that users might find unsettling ([apnews.com](https://apnews.com/article/a5937ca6778381907848e9a82d1131db#:~:text=Microsoft%20has%20introduced%20a%20new,a%20balanced%20approach%20to%20AI)). The goal is a useful AI persona that does not annoy or distract, in contrast to overly romanticized AI characters.