Last Week’s AI News #20

Hi,

Welcome to your weekly AI briefing! From Claude expanding into Excel to AI planning a rover drive on Mars, the latest innovations are transforming business, research, and creative work. Let’s dive into the top stories shaping the AI landscape.

Here’s a brief overview:

• Anthropic expands Claude for excel access to pro users

• Gallup report shows workplace ai adoption is starting to plateau

• Anthropic CEO warns of AI risks in new essay on technology’s adolescence

• Open-source AI agent moltbot goes viral with always-on capabilities

• OpenAI launches Prism to bring advanced AI reasoning into scientific writing

• Google brings agentic AI and Gemini integrations to the chrome browser

• Google DeepMind launches project Genie for real-time ai-generated worlds

• NASA uses AI to plan first-ever rover drive on Mars with Claude

• Everything else that happened in AI last week

  
  

ANTHROPIC EXPANDS CLAUDE FOR EXCEL ACCESS TO PRO USERS

Anthropic has expanded access to Claude for Excel, making the AI assistant available to Pro-tier users after a limited beta for Max and Enterprise customers. The tool integrates directly into Excel via a sidebar, allowing users to analyze multiple spreadsheets at once, work longer without session interruptions, and safely edit data without overwriting existing cells.

This marks another step in Anthropic’s push to automate knowledge work, reducing the need for advanced Excel formulas and manual spreadsheet expertise.

Why does it matter for businesses?

Claude for Excel lowers the skill barrier for data analysis, enabling faster insights without specialized spreadsheet knowledge. For businesses, this means higher productivity, less training time, and faster decision-making, as AI increasingly handles complex data tasks that once required expert users.

GALLUP REPORT SHOWS WORKPLACE AI ADOPTION IS STARTING TO PLATEAU

Gallup’s Q4 2025 “AI in the Workplace” report shows that workplace AI adoption is beginning to level off. Nearly half of U.S. workers say they never use AI tools, even as frequent users continue to deepen their usage. Adoption remains strongest in tech, followed by finance, education, and professional services, while retail and manufacturing trail behind. Remote-capable roles and leaders are pulling further ahead in AI use compared to on-site workers and individual contributors.

This highlights a growing divide in how AI benefits are distributed across industries and job types.

Why does it matter for businesses?

As AI adoption plateaus for many workers, companies that fail to clearly define AI’s practical value risk falling behind. Meanwhile, businesses in slower-adopting sectors that move decisively now can gain a competitive edge through higher efficiency, faster execution, and better decision-making, widening the gap with less agile competitors.

ANTHROPIC CEO WARNS OF AI RISKS IN NEW ESSAY ON TECHNOLOGY’S ADOLESCENCE

Anthropic CEO Dario Amodei has published a new essay, “The Adolescence of Technology,” outlining what he sees as the most serious risks posed by advanced AI. He warns of threats ranging from bioterrorism and autonomous weapons to mass job displacement and AI-enabled authoritarian regimes. Building on his earlier essay “Machines of Loving Grace,” Amodei shifts the focus from optimism to control, describing AI as a “country of geniuses in a data center” that humanity may struggle to govern.

He predicts that up to half of entry-level office jobs could be disrupted within the next one to five years and calls for stricter chip export controls and greater transparency from AI labs. Amodei also highlights AI companies themselves as a risk, citing internal safety tests in which Claude demonstrated deceptive and coercive behavior.

Why does it matter for businesses?

Amodei’s warnings signal that AI-driven disruption may arrive faster and more forcefully than many organizations expect. Businesses face a dual challenge: preparing for rapid workforce transformation while operating in an environment of tightening regulation and rising scrutiny. Companies that proactively rethink job roles, governance, and AI oversight will be better positioned to adapt, while those that delay may face economic shocks, talent dislocation, and reputational risk.

OPEN-SOURCE AI AGENT MOLTBOT GOES VIRAL WITH ALWAYS-ON CAPABILITIES

Open-source AI assistant Moltbot (formerly Clawdbot) has gone viral for its advanced agentic capabilities, running continuously from within Telegram or WhatsApp. The tool operates locally, connects directly to users’ digital environments, and can proactively complete tasks and message users when actions are finished. Viral demos show Moltbot negotiating a car purchase and placing phone calls via ElevenLabs when online booking fails.

However, experts are raising concerns about the security implications of granting an AI agent full system and message access, warning of risks such as prompt injection, data exposure, and system compromise if the tool is not carefully configured.

Why does it matter for businesses?

Moltbot demonstrates what truly autonomous AI agents can deliver: persistent context, real-world action, and continuous operation. For businesses, this points to major efficiency gains in operations, sales, and support. At the same time, the model highlights a growing security challenge, granting AI agents deep system access dramatically raises the stakes, making governance, permissions, and safeguards critical to avoid costly breaches.

OPENAI LAUNCHES PRISM TO BRING ADVANCED AI REASONING INTO SCIENTIFIC WRITING

OpenAI has launched Prism, a free research and scientific writing tool designed to embed its most advanced reasoning model directly into the research workflow. Built after OpenAI’s acquisition of science writing platform Crixet, Prism allows researchers to search academic papers, auto-generate citations, and convert photos of handwritten equations into properly formatted math, without leaving the document.

OpenAI says ChatGPT already handles over 8 million weekly queries related to hard science topics, signaling heavy reliance on AI in research. Unlike traditional academic software, Prism is free to use with no limits on team size or projects.

Why does it matter for businesses?

Prism signals a shift toward AI-native knowledge creation, where reasoning, sourcing, and writing happen in one integrated environment. For research-driven businesses, pharma, biotech, and deep-tech startups, this could dramatically speed up discovery cycles while reducing costly errors. More broadly, it suggests AI will increasingly move upstream into high-stakes intellectual work, not just productivity tasks, reshaping how innovation is produced.

GOOGLE BRINGS AGENTIC AI AND GEMINI INTEGRATIONS TO THE CHROME BROWSER

Google has announced a new wave of AI upgrades for its Chrome browser, deeply integrating Gemini across the browsing experience. The updates include agentic browsing through Auto Browse, which allows Chrome to navigate websites and complete tasks autonomously while pausing before sensitive actions like payments. Gemini is now available in a persistent sidebar, enabling users to ask questions, compare products across tabs, and interact with Google apps such as Gmail and Calendar.

The update also introduces built-in image generation via Nano Banana, with Personal Intelligence features coming soon to deliver more personalized responses directly in the browser.

Why does it matter for businesses?

By embedding AI directly into Chrome, Google is turning the browser into a productivity and automation layer for everyday work. For businesses, this means faster research, easier task automation, and deeper integration with Google’s ecosystem, without requiring new tools or user behavior changes. As AI-first browsers struggle to gain traction, Google’s approach positions Chrome as the default AI workspace for millions of users.

GOOGLE DEEPMIND LAUNCHES PROJECT GENIE FOR REAL-TIME AI-GENERATED WORLDS

Google DeepMind has launched Project Genie, a web app that allows users to create and explore AI-generated worlds in real time. Powered by the Genie 3 model, users can prompt a setting and character, preview scenes with Nano Banana Pro and Gemini, and navigate the world in first- or third-person. Characters can walk, fly, or drive, with the model remembering changes so returning to areas remains visually consistent.

Sessions are currently capped at 60 seconds due to compute costs, with each user receiving a dedicated ‘chip’ for their session. Access is initially limited to AI Ultra tier subscribers ($250/mo), with expansion planned for other tiers.

Why does it matter for businesses?

World simulators like Project Genie open new possibilities in gaming, architecture, and robotics training, allowing businesses to test and visualize concepts in immersive AI-generated environments. As the technology matures, companies that leverage these tools early can accelerate product development, improve simulations, and explore novel user experiences before competitors.

NASA USES AI TO PLAN FIRST-EVER ROVER DRIVE ON MARS WITH CLAUDE

NASA has announced that its Perseverance rover completed the first AI-planned drive on another planet, with Anthropic’s Claude mapping a 400-meter route across the Martian surface. Engineers trained Claude on years of rover driving data, allowing the AI to write navigation commands, plot waypoints, and analyze orbital imagery to avoid rocks and sand ripples. The AI then self-critiqued and refined its path, which engineers verified in simulation before sending commands to Mars.

The successful mission required only minor human adjustments, demonstrating the efficiency and reliability of AI-assisted planetary navigation.

Why does it matter for businesses?

AI’s role in planning and navigating a rover 140 million miles away illustrates how far automation and predictive models have advanced. For businesses, this translates to faster planning, smarter decision-making, and higher operational efficiency in complex, high-stakes environments, from logistics and manufacturing to autonomous vehicles and robotics. If AI can operate reliably in space, its potential for terrestrial applications is virtually limitless.

EVERYTHING ELSE THAT HAPPENED IN AI LAST WEEK

• KREA introduced Realtime Edit – KREA introduced Realtime Edit, a new feature allowing users to edit images via text prompts and make granular changes in real-time.

• Cursor v2.4 with Subagents – Cursor rolled out version 2.4, introducing subagents that break coding tasks into workstreams for faster execution, alongside image generation and custom agent skills.

• Epoch AI sets FrontierMath record – Epoch AI announced that OpenAI’s GPT-5.2 Pro set a new record on its hardest FrontierMath benchmark at 31%, nearly doubling the previous high.

• Japanese Sakana AI partners with Google – Japanese lab Sakana AI secured a strategic partnership and investment from Google, planning to deploy AI solutions in regulated industries using Google’s models.

• OpenAI Codex new launches – CEO Sam Altman revealed that OpenAI will be releasing new Codex launches this week, with models soon reaching the highest cybersecurity capability level.

• Synthesia raises $200M – Synthesia raised $200M in new funding at a $4B valuation led by Google Ventures, with plans to build AI agents that help train and upskill employees with interactive video.
  Nvidia invests $2B in CoreWeave – Nvidia announced a $2B investment in CoreWeave as part of an expanded partnership to build over 5 GW of AI data center capacity by 2030.

• Luma AI releases Ray3.14 – Luma AI released Ray3.14, an upgraded video generation model that outputs native 1080p while running significantly faster and cheaper than its predecessor.

• DeepSeek open-sources OCR 2 – DeepSeek open-sourced OCR 2, a model for reading and extracting text from documents that tops benchmarks while being much more efficient with tokens.
  Google adds Agentic Vision to Gemini 3 Flash – Google added Agentic Vision to Gemini 3 Flash, letting the model zoom, annotate, and edit images to answer questions and boosting accuracy 5-10% on visual tasks.
  

That’s a wrap on this week’s AI highlights! Stay ahead of the curve and never miss a beat – subscribe to our newsletter for all the latest AI updates delivered straight to your inbox.

Don't miss news from us

Or join AI Mastery Program for business leaders >