Google launches Gemini as Nvidia introduces Dynamic Context Parallelism

Key Takeaways

• Google Chrome introduced new AI features on January 28, 2026, including the Gemini AI assistant sidebar, Nano Banana for image generation/editing, and "Auto Browse" for multi-step tasks.
• The "Auto Browse" feature in Chrome is available for AI Pro and Ultra subscribers in the U.S. and requires manual confirmation for sensitive actions like purchases.
• NVIDIA launched Dynamic Context Parallelism on January 28, 2026, within its Megatron Core, speeding up large language model training by up to 1.48x.
• The UK government, with support from Amazon, Google, and Microsoft, announced free AI skills courses for 10 million adults and an "AI and Future of Work Unit" on January 28, 2026.
• DigitalOcean appointed Vinay Kumar as CPTO in January 2026 to aggressively scale its AI inference and core cloud platform.
• On January 27, 2026, the Doomsday Clock moved to 85 seconds to midnight, citing uncontrolled artificial intelligence as a major global threat.
• Researchers warned on January 22, 2026, about next-generation AI "swarms" that could infiltrate social media, mimic humans, and secretly influence public opinion.
• A report on January 28, 2026, found that nearly 70% of U.S. shoppers would use AI for routine tasks but trust banks and payment providers more than retailers for autonomous purchases.
• AI tools, such as Google's NotebookLM, offer speed but can "hallucinate" or misinterpret information, requiring human experts to ensure factual reliability for Social Purpose Organizations.
• Explore Asheville is using AI for "conversational commerce" and "answer engine optimization" to attract travelers and adapt to new search behaviors.

Google Chrome gets AI upgrades including Auto Browse

Google introduced new AI features for its Chrome browser on January 28, 2026. A new side panel keeps the Gemini AI assistant available across tabs, allowing users to multitask and connect with Google apps like Gmail and Calendar. Google also added Nano Banana, an AI image generation tool, to Chrome for in-browser image editing. The most advanced feature, "Auto Browse," is for AI Pro and Ultra subscribers, letting Gemini complete multi-step tasks like reordering items or filling out forms across various websites. This aims to make browsing more efficient by automating complex actions.

Chrome boosts AI with Gemini and Auto Browse features

On January 28, 2026, Google enhanced its Chrome browser with deeper Gemini AI integration and new agentic features. The Gemini sidebar now understands groups of tabs for better context, and it is available for Windows, macOS, and Chromebook Plus users. Chrome also includes Nano Banana for modifying images directly in the browser. The "Auto Browse" feature, for AI Pro and Ultra subscribers in the U.S., allows Gemini to perform multi-step tasks like shopping or filling forms, asking for user approval on sensitive actions. Google states that early testers used it for various tasks including scheduling and collecting tax documents.

Google brings AI image tools and assistant to Chrome

On January 28, 2026, Google added new AI capabilities to its Chrome browser for desktop users in the United States. The AI image generator and editing tool, Nano Banana, is now available directly in Chrome. Users can also open a side panel for an AI-powered assistant to help with various online tasks. Additionally, Google AI Pro and Ultra subscribers can use an "auto browse" feature that logs into websites, shops for items, and prepares social media posts. Users must still manually confirm purchases and approve social media drafts.

Chrome gets AI image tools and virtual assistant

Google is enhancing its Chrome browser with new artificial intelligence features. The AI image generator and editing tool, Nano Banana, is now available to logged-in Chrome users on desktop computers in the United States. A new side panel option also lets U.S. users access an AI-powered assistant for various online tasks. Google AI Pro and Ultra subscribers can use an "auto browse" function to log into websites, shop, and prepare social media posts, though users must manually complete purchases and approve posts.

Google adds more Gemini AI tools to Chrome

On January 28, 2026, Google announced more Gemini AI features for its Chrome browser. A new Chrome design includes a right-side panel that keeps the Gemini chatbot accessible while users browse. These features include the Nano Banana image generation tool and Google's "Personal Intelligence" feature. "Auto browse," an agentic AI tool, allows users to command the browser to perform various web tasks with a single prompt. AI Pro and Ultra subscribers in the U.S. will gain access to Auto browse, which can handle tasks like filtering apartments on Redfin.

Google Chrome introduces Auto Browse AI agent

Google has launched a new "Auto Browse" feature for Chrome, integrating more AI into the browser. Accessed through the Gemini sidebar, this agentic AI can take over browsing tasks. Charmaine D'Silva demonstrated how Auto Browse could reorder a jacket and find a discount coupon. The feature operates in its own tab, making "ghostly clicks" to complete tasks. Google warns users to be careful and take control if needed, as they are responsible for the AI's actions. Sensitive tasks like social media posting are currently restricted for security reasons.

Google launches Chrome Auto Browse AI for subscribers

Google is rolling out its "Auto Browse" AI agent for Chrome, expanding Gemini's capabilities. The AI is now easily accessible via a split-screen "Sidepanel" view and can connect to various Google services like Gmail and Google Flights. Chrome's Gemini can also use Nano Banana to edit images directly. Auto Browse, currently in preview, helps users with tedious tasks by opening new tabs and performing actions, marked with a sparkly AI icon. This feature is available to AI Pro and Ultra subscribers, with daily task limits. Google ensures security by requiring manual confirmation for purchases and adhering to privacy policies for streamed content.

Doomsday Clock moves closer to midnight due to AI and climate

On January 27, 2026, the Bulletin of the Atomic Scientists announced the Doomsday Clock is now set to 85 seconds to midnight, the closest it has ever been. This move from last year's 89 seconds reflects growing global threats. The scientists cited risks including nuclear war, the climate crisis, potential misuse of biotechnology, and uncontrolled artificial intelligence. They also noted escalating conflicts involving nuclear-armed countries like the Russia-Ukraine war and the India-Pakistan conflict. The group emphasizes that international trust and cooperation are essential to reverse this trend.

Doomsday Clock nears midnight due to global threats

On January 27, 2026, the Bulletin of the Atomic Scientists moved the Doomsday Clock to 85 seconds to midnight, indicating humanity is closer than ever to destruction. The group cited major threats including nuclear war, the climate crisis, and the uncontrolled use of artificial intelligence. They also highlighted increasing aggression among nations like Russia, China, and the U.S., and ongoing conflicts such as the Russia-Ukraine war. Daniel Holz, chair of the science and security board, stressed the need for international cooperation to address these existential risks and turn the clock back.

Shoppers trust banks more than retailers for AI purchases

A report on January 28, 2026, shows that while many U.S. shoppers are open to AI assistants for online tasks like grocery shopping, they trust banks and payment providers more than retailers for autonomous purchases. Nearly 70% of consumers would use AI for routine tasks, but their willingness to let AI complete transactions depends on the platform. Financial institutions have built decades of trust in security and oversight for automated buying. Only 16% of shoppers trust merchant-offered AI agents the most, preferring to use AI for browsing and then manually completing purchases themselves.

DigitalOcean boosts AI plans with new CPTO Vinay Kumar

In January 2026, DigitalOcean appointed Vinay Kumar as its Chief Product and Technology Officer to lead product strategy, cloud infrastructure, and security. Kumar, who helped build Oracle Cloud Infrastructure, will focus on scaling DigitalOcean's AI inference and core cloud platform. DigitalOcean aims to provide accessible and scalable infrastructure for AI model training and inference, offering specialized virtual machines and managed Kubernetes services. This hire signals an aggressive push into the AI infrastructure market, with a focus on ease of use and predictable pricing to attract businesses and developers.

NVIDIA speeds up AI training with Dynamic Context Parallelism

On January 28, 2026, NVIDIA introduced Dynamic Context Parallelism, or Dynamic-CP, within its Megatron Core. This new scheduling method aims to speed up large language model post-training and DiT pre-training by efficiently handling variable-length sequences. Real-world datasets often have sequences of different lengths, which can slow down training and waste resources. Dynamic-CP dynamically adjusts the context parallelism size for each microbatch, leading to up to a 1.48x speedup on real datasets. This approach helps overcome issues like GPU idling and communication overhead caused by imbalanced workloads.

AI swarms could mimic humans and invade social media

Researchers warn that next-generation AI "swarms" are poised to infiltrate social media, mimicking human behavior and potentially harassing real users. A commentary published on January 22 in First Monday describes these swarms as a new form of information warfare, capable of creating the illusion of authentic online movements. Coordinated by large language models, these sophisticated AI agents can adapt to online communities and target individuals with dissenting views. Iyad Rahwan and Janos Kiss explain that these swarms could secretly influence public opinion by exploiting human tendencies to follow the herd. The researchers note that these swarms will be difficult to detect, and their full impact remains unknown.

Explore Asheville uses AI to attract future travelers

On January 28, Explore Asheville announced it is using AI to adapt to changing travel search habits and attract more visitors. The organization is partnering with Visit Widget for "AI-powered conversational commerce," allowing travelers to discover and plan trips through conversation. Explore Asheville is also investing in "answer engine optimization" to ensure its information appears accurately in AI-generated search results. This involves structuring content and adding FAQs about local businesses and events. Vic Isley, president and CEO, emphasized the importance of ethical AI use to maintain an authentic community voice.

UK offers free AI courses with Amazon, Google, Microsoft

On January 28, 2026, the UK government announced free AI skills courses for all UK adults, with contributions from major tech companies like Amazon, Google, and Microsoft. This initiative aims to train 10 million people and make the UK the fastest G7 country to adopt AI. The government is also creating an "AI and Future of Work Unit" to research and advise on AI's effects on the economy and job markets. This project seeks to boost AI literacy and prepare the workforce for future technological changes.

AI offers speed but lacks stability for social groups

A report published on January 28, 2026, highlights a trade-off for Social Purpose Organizations using AI tools like Google's NotebookLM. While AI offers speed in creating presentations and summaries, it lacks stability and factual reliability. AI tools can "hallucinate" or misinterpret information, which is dangerous for SPOs that rely on trust and credibility. Experiments showed that AI translations can distort meaning, like changing "leadership" to "to rule" in Hindi, and alter infographic designs without explanation. To manage these risks, human experts are still needed to vet content, translations, and designs, which adds back the time and cost AI was meant to save.

Gen Z uses AI but also has concerns

A report from January 28, 2026, explores how Generation Z uses generative AI and why they have concerns about it. Young adults are often early adopters of new technologies, including AI. The article aims to understand their engagement with AI tools and the worries they experience despite their usage.

Sources

• Google Unleashes AI in Chrome with 'Auto Browse'
• Chrome takes on AI browsers with tighter Gemini integration, agentic features for autonomous tasks
• Google adds AI image generation to Chrome, side panel option for virtual assistant
• Google Adds AI Image Generation to Chrome Browser, Side Panel Option for Virtual Assistant
• Google brings more Gemini AI features to Chrome browser
• Google’s New Chrome ‘Auto Browse’ Agent Attempts to Roam the Web Without You
• Google begins rolling out Chrome's "Auto Browse" AI agent today
• Doomsday Clock at 85 seconds to midnight amid threats from climate crisis and AI
• 'Doomsday Clock' ticks closer to midnight amid threats from AI, climate change and nuclear war
• Shoppers Trust Banks Over Retailers to Let AI Buy
• Is DigitalOcean (DOCN) Quietly Recasting Its AI Infrastructure Ambitions With a New CPTO Hire?
• Speeding Up Variable-Length Training with Dynamic Context Parallelism and NVIDIA Megatron Core
• Next-generation AI 'swarms' will invade social media by mimicking human behavior and harassing real users, researchers warn
• Explore Asheville looks to AI to adapt to changing travel search habits
• Amazon, Google, Microsoft among providers bringing free AI courses to all UK adults | MLex | Specialist news and analysis on legal risk and regulation
• Speed without stability: The AI trade-off facing social purpose organisations
• How Gen Z Uses Gen AI—and Why It Worries Them