Good morning, { AI enthusiasts }. The U.S. government just told an AI company to remove the safeguards keeping its model from getting used in autonomous weapons and mass surveillance… Or potentially be forced to legally comply.
The Pentagon’s ultimatum to Anthropic is a scary one, and the way it plays out could set the precedent for the way probably the most powerful AI systems on the planet find yourself getting used by governments, with some dystopian potential outcomes.
In today’s AI rundown:
-
Pentagon hits Claude with ultimatum over guardrails
-
Latest AI learns any computer task by watching videos
-
Translate videos into any language with AI
-
Cowork gets agentic tools across departments
-
4 recent AI tools, community workflows, and more
LATEST DEVELOPMENTS
ANTHROPIC & THE PENTAGON
Image source: Full Metal Jacket / The Rundown
The Rundown: The Pentagon and Defense Secretary Pete Hegseth just delivered a Friday ultimatum to Anthropic CEO Dario Amodei: remove Claude’s military safeguards, or face contract termination, government blacklisting, and compelled compliance.
The main points:
-
Amodei has refused to budge on two Claude uses — autonomous weapons with out a human within the loop, and bulk surveillance of Americans.
-
Axios reports Hegseth gave three selections: Agree, kill the $200M cope with a supply chain risk label, or be forced to comply via the Defense Production Act.
-
Claude was the primary model contained in the Pentagon’s classified networks, with xAI’s Grok now landing a deal after agreeing to “all lawful purposes” use cases.
-
The Pentagon can be reportedly fast-tracking OpenAI and Google for classified access, giving it additional options if Anthropic walks.
Why it matters: AI is becoming a pillar of military power, but strong-arming labs into dropping safety limits sets a bleak precedent. If ‘wartime’ legal threats are enough to strip guardrails, the query of who will actually draw the road on dystopian AI uses like autonomous weapons and surveillance quickly gets a really unsettling answer.
TOGETHER WITH SLACK FROM SALESFORCE
The Rundown: Slackbot is a context-aware AI agent built directly into Slack — understanding your conversations, files, and workflows to deliver what you would like, right if you need it, with zero setup.
Watch this 2-minute demo to see how Slackbot:
-
Makes your entire workspace searchable (docs, convos, apps)
-
Enhances every teammate with role-specific automations
-
Learns your project and preferences over time for even smarter outputs
-
Synthesizes what you would like immediately, respecting permissions and using only what you’ll be able to already see
STANDARD INTELLIGENCE
Image source: Standard Intelligence
The Rundown: Standard Intelligence introduced FDM-1, a ‘computer motion’ model that learns to operate computers by watching video — already showing it may possibly do CAD modeling, find software bugs, and drive an actual automobile through San Francisco.
The main points:
-
FDM-1 is trained on 11M hours of screen footage, 550,000x (!) the biggest open dataset, with an AI that reverse-engineers what actions produced each frame.
-
The model can watch and follow together with nearly two hours of continuous screen activity without delay, processing 50x the visual context of existing models.
-
FDM-1 Demos range from constructing gears in Blender to driving an actual automobile via arrow keys and live data feeds with under an hour of coaching data.
Why it matters: Language models learned how we write from the web’s text, and FDM-1 is now attempting to learn the way we work and operate from the web’s video. By enabling far more of the world’s video to be ingested as training data with higher retention, the ceiling for what computer-use agents can just do jumped dramatically.
AI TRAINING
The Rundown: On this guide, you’ll learn methods to use HeyGen to translate your videos into any language. When you’re running paid or organic video content to a world audience, it is a powerful, easy-to-use solution to expand your reach across the globe.
Step-by-step:
-
First, perform a little research in your Meta/Google ads accounts on your top 3 non-English markets
-
We’re using a free HeyGen account, which allows three translations. Keep your video under 3 minutes — a single-speaker talking head format works best
-
Click Translate in HeyGen and upload the clip or provide a YT/Google Drive link. Next, select the required language without changing the default settings
-
In 5–10 minutes, your video can be ready so that you can download or share via link
Pro tip: If the standard looks low, a paid plan will let you export in the unique quality.
PRESENTED BY LAMBDA
The Rundown: Enterprises stuck on large, closed models are paying more for less control. Lambda and Oumi have partnered to vary that — delivering end-to-end custom model development and deployment in hours with higher accuracy, lower latency, and stronger privacy in-built.
Their platform delivers:
-
Automated test set creation, evaluation, and fine-tuning end-to-end
-
Deployment on secure, single-tenant production infrastructure
-
Proven results, with 70% cost savings and 20% quality gains in a single healthcare deployment
ANTHROPIC
Image source: Anthropic
The Rundown: Anthropic released a serious update to its Cowork agentic platform with recent department-specific AI agents, private plugin stores, and recent connectors for Gmail, DocuSign, and more — escalating the enterprise agent war with OAI’s Frontier.
The main points:
-
Latest pre-built agents cover 10 departments out of the gate, starting from HR and engineering to banking, equity research, and wealth management.
-
Latest connectors include Google Workspace, DocuSign, FactSet, and Harvey — with added plugins for partners like Slack by Salesforce, S&P Global, and LSEG.
-
Firms can construct private agent stores and push custom AI agents to specific teams, with admin controls to limit and assign access.
-
A brand new research preview also allows Claude to hop between Excel and PowerPoint, crunching data in a single and constructing a full deck in the opposite.
Why it matters: Cowork spooked SaaS stocks at launch as a mere research preview, and now it ships with agents for 10 departments and connectors to the tools those teams already use. Anthropic is bolting on a brand new sector with every update — and if this pace holds, all the knowledge economy starts to appear like one big Claude wrapper.
QUICK HITS
-
💼 Claude Cowork – Anthropic’s agentic platform, with recent team plugins
-
🤖 Custom Agents – Notion’s always-on AI agents for automating workflows
-
🎨 Seedream 5.0 Lite – ByteDance’s upgraded AI image model
-
🎆 Reve v1.5 – Reve’s recent text-to-image model with 4k resolution outputs
Explore the long run of AI-driven dev at Sonar Summit. Register here for the free, virtual event on March third.*
Anthropic launched Distant Control for Claude Code, letting users easily hand off running terminal tasks to their phone or browser.
Google Labs acquired AI music platform ProducerAI, plugging it into DeepMind’s Lyria 3 model to let creators generate full tracks and custom instruments from text prompts.
Inception Labs launched Mercury 2, a diffusion-based reasoning model clocking over 1,000 tokens/sec, tripling the speed of its closest competitor in the identical price tier.
Meta and chipmaker AMD announced a brand new multi-year deal for as much as 6GW of GPUs, the tech giant’s biggest move yet to interrupt free from Nvidia-only infrastructure.
OpenAI hired Arvind KC as its recent Chief People Officer, tapping a veteran of Roblox, Google, Palantir, and Meta to assist proceed scaling the AI giant.
COMMUNITY
Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.
Today’s workflow comes from reader Adam P. in England:
“We’re two part-timers, plus me, a founding father of 30 years. We were losing ground to greater competitors with big marketing teams. Our search engine optimization had stalled, business was flat.
We fed our web data and competitor information into AI. It ran SWOT comparisons, highlighted content gaps, and showed where we were being outranked. We used it to restructure our website and tighten our CRM workflows. Traffic is now growing again. Enquiries are stronger. We’re back in expansion mode.”
How do you employ AI? Tell us here.
-
Read our last AI newsletter: Anthropic calls out China’s AI copycats
-
Read our last Tech newsletter: Apple goes big on ‘Made in America’
-
Read our last Robotics newsletter: Figure’s 24/7 humanoid staff
-
Today’s AI tool guide: Translate your videos into any language with AI
-
RSVP to today’s workshop @ 2PM EST: Agentic Workflows Bootcamp pt. 3
That is it for today!
- ⭐️⭐️⭐️⭐️⭐️ Nailed it
- ⭐️⭐️⭐️ Average
- ⭐️ Fail
Login or Subscribe to participate
See you soon,
