Good morning, AI enthusiasts. Apple hasn’t been able to get Siri right, but its smallest computer might be the one actually powering the AI agent era.

Perplexity just spun out its Computer agentic system with a ‘Personal’ local option, pitching itself as the safer alternative to OpenClaw — and likely ramping up the demand for Apple’s small but mighty machine even more in the process.

Reminder: Our next live workshop is today at 2 PM EST — join pt. 1 of our Intro to Vibe Coding Workshop and learn how to build reusable, working software for your specific tasks without any technical experience needed. RSVP here.

In today’s AI rundown:

• Perplexity turns Mac mini into a 24/7 AI agent

• Musk revives Macrohard as a joint xAI-Tesla project

• Create agentic workflows in Google Workspace

• Anthropic Institute to document AI’s disruption

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

PERPLEXITY

💻 Perplexity turns Mac mini into a 24/7 AI agent

Image source: Perplexity

The Rundown: Perplexity just debuted Personal Computer, a new local version of its Computer AI agent system that runs on a dedicated Mac mini — positioning itself as a more secure, controlled rival to the viral OpenClaw.

The details:

• The agent gives Perplexity's Comet assistant persistent local access to files, apps, and sessions on a Mac mini, able to be managed from anywhere.

• Perplexity frames it as a safer OpenClaw, with safeguards like tracked activity, sign-off for sensitive tasks, and a ‘kill switch’ to shut the system down.

• Perplexity Computer launched in late February, providing a cloud-based agentic system that orchestrates different models at once to complete tasks.

• Perplexity Max subscribers get early access via a waitlist, with the company saying it will provide “support and resources” for the initial cohort of users.

• Perplexity also released Computer to enterprise, tapping into 20 models and 400+ app connections — along with a Slack integration for team workflows.

Why it matters: Everyone knocks Apple for its AI mishaps, but the Mac mini is unintentionally becoming the default hardware for the AI agent era. Between OpenClaw, Perplexity's Personal Computer, and the wave of offshoots, always-on local agents are getting safer and easier to set up — and soon everyone's going to have one.

TOGETHER WITH SLACK FROM SALESFORCE

♥️ Why Microsoft customers love Slack

The Rundown: How do top teams use Microsoft 365? They integrate it with Slack to create an all-in-one work operating system, connecting every tool, automating workflows, and putting AI at the center.

Watch this demo to see how 150,000+  Microsoft customers maximize existing tech investments with Slack, including:

• Best in class integrations with Office, SharePoint, Teams, etc.

• AI-powered automation and no-code search

• An open ecosystem with 2,600+ app connections

Watch and learn more.

XAI & TESLA

🤖 Musk revives Macrohard as a joint xAI-Tesla project

Image source: Lovart / Elon Musk on X

The Rundown: Elon Musk pushed back on reports that its AI software initiative Macrohard had stalled, revealing xAI's project is merging with Tesla's ‘Digital Optimus’ AI agent into a system he claims can "emulate the function of entire companies."

The details:

• The system will pair Grok with a ‘Digital Optimus’ agent, processing live screen video and inputs and borrowing techniques from Tesla's FSD tech.

• Musk says it'll run on Tesla's $650 AI4 chip paired with xAI's Nvidia servers — and calls it "the only real-time smart AI system" available.

• The post came after Business Insider reported that 20+ Macrohard engineers had left or shifted roles, with a 600-person data project also on pause.

• xAI merged with SpaceX in February, but has suffered from a wave of employee exits over the last month, including several co-founders.

Why it matters: "Emulate the function of entire companies" is a big claim, but Musk has the pieces for it — custom chips, FSD-trained video processing, and Grok's reasoning in one stack. Macrohard could end up as the most vertically integrated agent play on the market… But his grand visions often take longer than expected to develop.

AI TRAINING

🛠️ Create agentic workflows in Google Workspace

The Rundown: In this guide, you will learn how to create an agent in your Google Workspace that reads form submissions, tracks them in a spreadsheet, and then emails you action items using Google’s new AI automation tool called Workspace Studio.

Step-by-step:

1. Go to Workspace Studio (request access if unavailable), create a new Flow, select “When a form response comes in” as the starter, and pick your form

2. Add a new step and select Summarize > Content from previous steps > Variable > Form response. In the prompt box, give the form’s purpose

3. Add a Decide step with the prompt “Read this summary [summary variable] and decide if it meets [your criteria for escalation]”

4. Click Add substep > Notify me by email and add the AI summary in the message body along with any other form variables

Pro tip: You can use this workflow to triage inbound leads, handle internal troubleshooting, onboard clients, and more.

PRESENTED BY TELY AI

💬 Market leaders get leads from ChatGPT and Google

The Rundown: Your buyers are asking AI questions — and AI is answering with your competitors, not you. Tely makes AI like ChatGPT, Google, and Claude recommend your business instead.

With Tely AI, you can:

• Get recommended in ChatGPT, Google, Perplexity, and Claude in as little as 1 week

• Fully hands-off: no writers, no agencies, no managing content

• Costs less than hiring freelancers or maintaining a marketing team

• Ideal for niche industries where expertise matters

Get leads from Google and ChatGPT on autopilot.

ANTHROPIC

🏛️ Anthropic Institute to document AI’s disruption

Image source: Lovart / The Rundown

The Rundown: Anthropic unveiled the Anthropic Institute, a new group combining three teams under co-founder Jack Clark to study AI’s societal impacts, launching amid the company's legal battle with the Pentagon over its blacklisting as a supply-chain risk.

The details:

• The ~30-person team will merge Anthropic's Frontier Red Team, Societal Impacts, and economics research groups, with plans to double staff yearly.

• The Institute plans to share learnings from building frontier models with the public, while engaging workers and industries facing AI displacement head-on.

• Founding hires include ex-DeepMind researcher Matt Botvinick, economist Anton Korinek, and Zoe Hitzig, who resigned from OAI over ads in ChatGPT.

Why it matters: Anthropic has not been shy about banging the drum on AI’s coming disruption, and now it has a whole think tank devoted to it. If a powerful AGI-level system really does arrive this year (and some may argue they’re already here), having an institute already studying its fallout may turn out to be one of the smarter bets in AI.

QUICK HITS

🛠️ Trending AI Tools

• 🗣️Unwrap Customer Intelligence - Turn unstructured customer feedback into data-backed insights that inform your product roadmap*

• ⚙️ Agent 4 - Replit's coding agent with infinite design canvas, team workflows

• 🧠 Nemotron 3 Super - Nvidia's 120B reasoning AI with 1M token context

• 🎧 TADA - Hume AI's open-source TTS AI with expressive and reliable outputs

*Sponsored Listing

📰 Everything else in AI today

Replit raised $400M at a $9B valuation, while dropping Agent 4, a coding agent that ships 10x faster with parallel agents, deeper collaboration, and broader build options.

Microsoft filed an amicus brief in support of Anthropic in its lawsuit against the Pentagon's supply chain blacklist, calling for a restraining order on the ban.

Amazon imposed a 90-day code safety reset after AI changes led to outages that cost 6.3M lost orders in one day, now requiring dual sign-offs for crucial deployments.

NVIDIA released Nemotron 3 Super, an open-source 120B reasoning model built for multi-agent workflows with a 1M-token context window and 5x faster speeds.

Cloudflare introduced a /crawl API endpoint that scrapes entire websites in one call, in a notable pivot from the company known for selling anti-bot protection.

Amazon launched Health AI, a free agentic assistant that can read medical records, book appointments, and manage prescriptions, with five free visits for Prime members.

COMMUNITY

🤝 Community AI workflows

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 Jigar P. in Canada:

"I use AI as a learning partner to support my son’s communication and language development. He is on the Autism spectrum. I generate tailored WH-question exercises (who, what, where, when, why) so he can practice understanding and answering.

I also use AI to design vocal and speech exercises that help reduce echolalia by encouraging him to move from repeating the question to giving his own response. Instead of relying on generic worksheets, I create learning materials that adapt to what he finds challenging and what keeps him engaged. AI helps me produce new examples, practice prompts, and role-play scenarios, so the practice stays fresh and consistent."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Yann LeCun’s $1B bet against LLMs

• Read our last Tech newsletter: Air taxis to fly in 26 states

• Read our last Robotics newsletter: DJI pays $30k for robot vacuum hack

• Today’s AI tool guide: Create agentic workflows in Google Workspace

• RSVP to our next workshop today @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, AI enthusiasts. Few people in AI have been louder about LLMs being a dead end than Yann LeCun. Even fewer have a Turing Award and a billion dollars to do something about it.

His new Advanced Machine Intelligence just launched with over $1B in funding to build what he believes LLMs never can: AI that actually understands the real world.

In today’s AI rundown:

• LeCun's anti-LLM startup opens with $1B

• Meta acquires AI agent social media platform

• Replicate ChatGPT Pulse on the $20 plan

• Murati lands Nvidia deal for Thinking Machines

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

AMI LABS

💰 LeCun's anti-LLM startup opens with $1B

Image source: Advanced Machine Intelligence

The Rundown: Ex-Meta Chief Scientist Yann LeCun’s Advanced Machine Intelligence just emerged with a $1.03B seed round, with the Turing Award winner betting on a world model approach to AI over the LLM approach he’s been railing against for years.

The details:

• LeCun left Meta in November after 12 years with FAIR, telling Mark Zuckerberg he could build world models “faster, cheaper, and better” on his own.

• AMI's systems hope to simulate how the physical world works with persistent memory, targeting manufacturing, robotics, wearables, healthcare, and more.

• The round values the company at $3.5B, with Nvidia, Samsung, Bezos Expeditions, Eric Schmidt, and Mark Cuban among the backers.

• LeCun chose Paris for AMI's headquarters, calling Silicon Valley 'LLM-pilled' —with additional hubs in New York, Montreal, and Singapore.

Why it matters: That is quite the seed round, and LeCun has quickly landed on his feet after leaving the FAIR research team he called home for over a decade. The outspoken research scientist has gone against the LLM flow throughout the AI boom. Now, he finally has a major war chest and the freedom to work on his vision of world models.

TOGETHER WITH LAMBDA

⚡ Cut your AI training costs by 25% or more

The Rundown: Most large-scale AI training runs use less than half the computing power they're paying for. Lambda's team found the root causes and built a reproducible framework that boosted efficiency by over 25%—without changing the model itself.

Lambda’s whitepaper shows you how to address:

• Memory inefficiencies silently inflating your costs

• Training configurations that aren't making full use of your hardware

• Bottlenecks that slow down GPU communication

Get the guide.

META

🦞 Meta acquires viral AI agent social media platform

Image source: Moltbook

The Rundown: Meta acqui-hired the creators of Moltbook, the viral vibe-coded social forum for AI agents that went viral alongside OpenClaw — folding the duo into its Superintelligence Labs team, weeks after OpenAI hired OpenClaw’s Peter Steinberger.

The details:

• Co-creator Matt Schlicht launched Moltbook in late January as a weekend project, building most of it with his OpenClaw bot named ‘Clawd Clawderberg’.

• Mark Zuckerberg had reportedly tried OpenClaw and courted Steinberger first, but lost out after the developer joined OpenAI in February.

• The platform has 2.8M registered bots with nearly 200K verified to real people, forming what Meta calls an 'always-on directory' for agent coordination.

• Posts about bot religions and anti-human manifestos went viral, though researchers found security holes that let humans easily pose as agents.

Why it matters: Meta's feeds are already filling up with AI creators and bot content, but it's been a messy, backlash-heavy rollout. Moltbook offers something different — a verified agent layer where bots operate in the open, not disguised as humans. How Zuck and co. incorporate the concept will be an interesting angle to watch.

AI TRAINING

🧠 Replicate ChatGPT Pulse on the $20 plan

The Rundown: In this guide, you will learn how to replicate ChatGPT Pulse (from the $200 Pro plan) on your $20 Plus plan, as well as discover how to use the lesser-known ChatGPT feature.

Step-by-step:

1. Pulse proactively messages you each day with news and suggestions based on your chat history. ChatGPT also has a feature called “tasks”

2. Open ChatGPT on the web, desktop, or mobile and prompt: “Create a daily, recurring task that briefs me on the daily stock market moves at 5 PM”

3. To edit/delete tasks, go to Settings > Notifications > Tasks > Manage Tasks. Turn on push notifications for tasks and for the app on your phone

4. You should start getting daily, recurring briefs in the chat where you set up the task

Pro tip: You can have up to 10 active tasks at once, so we recommend setting up briefs on multiple topics or recurring tasks on connected services like Gmail or your Calendar.

PRESENTED BY UNWRAP

💬 Powerful insights for powerful brands

The Rundown: Unwrap’s customer intelligence platform brings all your customer feedback (surveys, reviews, support tickets, social comments, etc.) into a single view, then uses AI + NLP to surface the most actionable insights and deliver them straight to your inbox.

With Unwrap, you get:

• A clear, single view of customer sentiment

• The ability to ask questions about your feedback using natural language with Assistant

• Real-time alerts from your feedback as they arise

• Tools trusted at scale by Oura, Stripe, Clay, DoorDash, lululemon, Southwest Airlines, and others

Exclusive for The Rundown subscribers, connect with Unwrap for a free trial.

THINKING MACHINES & NVIDIA

💰 Murati lands Nvidia deal for Thinking Machines

Image source: Thinking Machines

The Rundown: Mira Murati’s Thinking Machines Labs just secured a multiyear deal with Nvidia for at least a gigawatt of compute, giving the former OpenAI exec’s year-old startup the kind of AI infrastructure typically reserved for the largest labs.

The details:

• Murati was OpenAI's CTO and briefly its interim CEO before leaving to found Thinking Machines, which pulled in $2B in funding at a $10B valuation last year.

• The multiyear deal puts at least a GW of Nvidia's next-gen Vera Rubin systems behind TML's frontier model training, with deployment targeted for early 2027.

• Nvidia also added undisclosed new capital on top of its existing stake from the $2B seed round, though neither company shared the size.

• TML has one product live, Tinker, a fine-tuning API for enterprises, but the gigawatt commitment signals a move toward creating their own models.

Why it matters: The exodus of TML employees and co-founders moving back to OpenAI in January looked like a death blow for a startup that had made little noise since launching. But this Nvidia partnership is a loud response and a clear sign that Murati has bigger ambitions — regardless of who is sticking around with her.

QUICK HITS

🛠️ Trending AI Tools

• 🤖 Thenvoi - Connect any agent, any framework with a powerful multi-peer agentic communication mesh*

• ⚙️ Code Review - Anthropic's multi-agent PR review system for Claude Code

• 🧠 Gemini Embedding 2 - Google’s AI to search across text, images, video, audio

• 💻 Interactive Learning - ChatGPT’s visual math and science modules

*Sponsored Listing

📰 Everything else in AI today

Google released Gemini Embedding 2 in public preview, its first AI model that can understand and search across text, images, video, and audio in a single system.

OpenAI introduced interactive visual modules to ChatGPT for 70+ math and science concepts, letting users tweak variables and watch formulas respond in real time.

Hume AI opened TADA, a speech generation AI that locks text & audio in sync to cut hallucinations, coming in 5x faster than rivals and light enough for on-device use.

Google upgraded Gemini across its productivity suite with the ability to draft docs, build sheets, and create presentations by pulling context from files, inbox, and the web.

Amazon secured a preliminary injunction against Perplexity’s Comet browser, barring the AI agent from buying products through Amazon accounts on behalf of users.

Nvidia is preparing to launch NemoClaw, an open platform for enterprises to run AI agents across any hardware, with early pitches out to Salesforce, Google, and others.

COMMUNITY

🤝 Community AI workflows

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 Sahar R. in St. Louis, MO:

"I'm a philosophy professor, and I integrated AI in nearly every step of generating a new course (Business Ethics) — from finding real-world cases posted about on credible news outlets, open-access case law, open educational resources as class materials, designing discussion board questions, building fair rubrics, and interactive in-class activities for the students.

Augmenting my course development with AI enhanced my course, the student experience, my effort and time, and my work became fun again."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Anthropic takes U.S. government to court

• Read our last Tech newsletter: Air taxis to fly in 26 states

• Read our last Robotics newsletter: DJI pays $30K for robot vacuum hack

• Today’s AI tool guide: Replicate ChatGPT Pulse on the $20 plan

• RSVP to our next workshop Thursday @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer—the humans behind The Rundown

Good morning, tech enthusiasts. This summer, electric air taxis from startups like Archer, Joby, and Beta will hit the skies in federally backed trials, shuttling passengers and cargo across 26 states.

The results could show whether urban air mobility is finally ready for liftoff, or still years away from escaping the hype cycle.

In today’s tech rundown:

• Electric air taxis to take flight in 26 states

• Apple preps high-end ‘Ultra’ line

• Bluesky CEO Jay Graber steps down

• Palmer Luckey’s gaming startup seeks $1B valuation

• Quick hits on other tech news

LATEST DEVELOPMENTS

EVTOL

🚀 Electric air taxis to take flight in 26 states

Image source: Beta Technologies

The Rundown: This summer, Americans in 26 states will get their first taste of “flying cars,” as federally backed pilot programs send electric air taxis from startups like Archer, Joby, and Beta skimming over city traffic under tightly controlled FAA trials.

The details:

• Startups including Archer, Joby, Beta, Wisk, Electra, and Reliable Robotics will run passenger, cargo, and emergency flights with state and local partners.

• The three-year program, created by a Trump executive order, lets these aircraft operate under tight supervision before full FAA type certification.

• Officials say the aircraft will be quieter, cleaner, and ultimately cheaper than helicopters, and aimed at cutting congestion and linking smaller cities.

• Beta CEO Kyle Clark said the program will let Beta start flying a year early; its shares jumped nearly 12% Monday, with Archer and Joby stock rising too.

Why it matters: Partnering with state and local governments, flying car startups get a chance to gather real‑world data across dozens of operation scenarios, from Manhattan heliports to routes over the Gulf Coast, access that could boost stock prices and validate their business models — or reveal just how far the tech still has to go.

APPLE

🍎 Apple preps high-end ‘Ultra’ line

Image source: Ideogram / The Rundown

The Rundown: Apple is reportedly getting ready to seriously crank up the luxury dial, prepping a $2,000‑ish foldable iPhone, a touchscreen MacBook Pro, and camera‑toting, AI‑smartened AirPods aimed squarely at the ultra‑high‑end crowd.

The details:

• Bloomberg’s Mark Gurman reports that at least three ‘Ultra’-tier devices are in the pipeline for 2026, including a foldable iPhone with a large inner display.

• The Ultra tier is built around a $2,000‑class foldable iPhone that sits well above today’s Pro models, as well as a higher-priced MacBook Pro.

• Apple is also said to be working on new AirPods with built‑in computer‑vision cameras that can feed visual “intelligence” straight into Siri

• The new MacBook is expected to pair the first touchscreen ever on a Mac with an OLED panel, at a price beyond today’s M5 Pro and M5 Max MacBook Pro.

Why it matters: The budget MacBook Neo may have grabbed headlines, but Gurman says Apple’s real play this year is at the other end of the price spectrum. A foldable iPhone, AI-camera AirPods, and a touchscreen OLED MacBook are all expected before year’s end — and none of them will be cheap.

BLUESKY

🪜 Bluesky CEO Jay Graber steps down

Image source: Wikimedia Commons

The Rundown: Bluesky CEO Jay Graber is stepping down to hand longtime tech operator Toni Schneider the reins — just as the Jack Dorsey-founded Twitter alternative finally hits its stride at 40M users, and its first real growing pains.

The details:

• Venture capitalist and ex-Automattic CEO Toni Schneider is taking over as interim CEO while Bluesky’s board searches for a permanent replacement.

• Graber says Bluesky has matured to the point where it needs a seasoned operator focused on scaling and execution.

• Bluesky has grown into a decentralized social network with more than 40M users and an ecosystem of over 500 third-party apps.

• The shake-up lands as Bluesky runs into new state age‑verification rules, forcing it to cut off Mississippi and start checking users’ ages in other states.

Why it matters: Graber became CEO of Bluesky in 2021, as it spun off from Twitter into its own company. As she shifts back toward Bluesky’s underlying AT Protocol — the open social framework she has worked on since the project’s early days —Schneider will have to face a growing wave of age-assurance regulation.

PALMER LUCKEY

🕹️ Palmer Luckey’s gaming startup seeks $1B valuation

Image source: Wikimedia Commons

The Rundown: Palmer Luckey, the Oculus co-founder turned defense-tech billionaire, is now reportedly chasing a $1B valuation for his retro gaming startup ModRetro, betting that there’s serious money in high-end nostalgia hardware.

The details:

• ModRetro’s Chromatic is a handheld built in the mold of the original Game Boy, with support for original GB and GBC cartridges. Price starts at $199.99.

• Luckey has described the Chromatic as the product of “hundreds of irrational decisions” intended to create a worthy tribute to the original Game Boy.

• The Financial Times reports ModRetro is already developing additional hardware, including a device meant to replicate the Nintendo 64.

• Luckey’s defense startup Anduril, whose autonomous weapons vision has been embraced by Trump, is also seeking a new funding round at a $60B valuation.

Why it matters: Luckey is testing whether premium nostalgia gadgets can earn a unicorn valuation in a handheld market already crowded with Analogue, Anbernic, AyaNeo, and a dozen retro rivals. The same engineer perfecting Game Boy clones is also racing firms like Palantir and Shield AI to automate warfare.

QUICK HITS

📰 Everything else in tech today

Meta AI chief Yann LeCun’s new startup, AMI, raised about $1B to build world-model systems that understand and reason about the physical world.

Apple reportedly delayed the launch of its long‑planned J490 smart home display because the overhauled Siri and on‑device AI assistant are still not ready.

AT&T said it plans to pour more than $250B into U.S. network infrastructure over the next five years.

Lightspeed Venture Partners and Andreessen Horowitz are backing Nexthop AI, a fast-growing AI data center supplier now valued at about $4.2B.

Uber is rolling out its “Women Drivers” feature nationwide in the U.S., allowing women and teens to request female drivers, expanding a safety-focused pilot.

XPRIZE founder Peter Diamandis is offering $3.5M for upbeat sci‑fi films that cast AI as the hero, not the villain.

Apple now produces 25% of all iPhones in India — around 55M units produced in 2025 — as it accelerates a shift away from China to dodge US tariffs.

Chinese tech giant Xiaomi is reportedly exploring vehicle-integrated solar tech to add photovoltaic panels to future EVs and extend their driving range.

Samsung told CNBC that its first AI smart glasses, featuring an eye-level camera that connects to a smartphone for AI processing, are planned for launch later this year.

Startup Cortical Labs is building test data centers in Melbourne and Singapore that swap server racks for “biocomputers” made from lab‑grown human brain cells.

OSHA is investigating the death of a 61-year-old contractor who was fatally pinned between a tractor-trailer and a loading dock at a Rivian warehouse in Illinois.

Bezos-backed EV startup Slate Auto replaced CEO Christine Barman with ex-Amazon VP Peter Faricy, just months before the launch of its first electric truck.

Peter Thiel’s Founders Fund is reportedly close to closing its fourth growth fund with about $6B in commitments.

Nintendo is suing the U.S. government to claw back Trump-era import tariffs that courts have ruled unlawful.

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Anthropic takes U.S. government to court

• Read our last Tech newsletter: Meta sued over Ray-Ban privacy

• Read our last Robotics newsletter: DJI pays $30k for robot vacuum hack

• Today’s AI tool guide: Auto-generate video from writing with Manus

• RSVP to our next workshop Thursday @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team

Good morning, AI enthusiasts. A few months ago, Anthropic was one of the Pentagon's most prominent AI partners. Now it's suing the government in federal court.

With two new lawsuits challenging the “supply chain risk” label and White House directive to cut ties, Anthropic is arguing it’s being retaliated against for speaking up on safety — and now 30+ employees from OpenAI and Google are lining up behind them.

In today’s AI rundown:

• Anthropic takes U.S. government to court

• Microsoft’s Claude-powered Copilot Cowork

• Auto-generate videos from content with Manus

• a16z releases new consumer AI Top 100

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

ANTHROPIC

⚖️ Anthropic takes U.S. government to court

Image source: Lovart / The Rundown

The Rundown: Anthropic fired back at the Trump administration with two lawsuits, challenging the Pentagon's 'supply chain risk' label and White House directive for all federal agencies to drop Claude — a move it calls punishment for its AI safety positions.

The details:

• Anthropic filed lawsuits in two separate courts, asking judges to throw out the blacklist label and block the government from forcing agencies to cut ties.

• The suits argue the ‘supply chain risk’ label was designed to counter foreign adversary threats, not punish a U.S company over policy disagreements.

• They also claim the Pentagon violated free speech rights by retaliating for publicly advocating AI safety limits on weapons and surveillance.

• 30+ OAI and Google staffers signed a legal brief backing Anthropic's Pentagon lawsuit, warning that the blacklisting threatens U.S. AI leadership.

Why it matters: Whatever side you take on who controls AI in warfare, the federal response (Hegseth's blacklist, Trump's Truth Social posts) has looked a lot like retaliation. Win or lose, the case could decide whether the gov. can move against a domestic company for speaking up on safety — a precedent all labs will be watching.

TOGETHER WITH TELEPORT

🤐 Your AI agents shouldn’t share secrets

The Rundown: Autonomous agents can’t rely on API keys or long-lived credentials anymore. Teleport’s Agentic Identity Framework replaces static secrets with verifiable, cryptographic identity for machine actors at scale.

With Teleport, you get:

• Zero standing privileges for autonomous agents

• Ephemeral, certified machine identities

• Policy-based access control

• Reliable, auditable workflow orchestration

If agents can act, they must be identified. Explore the framework.

MICROSOFT

🤖 Microsoft’s Claude-powered Copilot Cowork

Image source: Microsoft

The Rundown: Microsoft just introduced Copilot Cowork, a new M365 feature built on Anthropic's Claude system that runs tasks in the background across apps — launching alongside a $99 enterprise bundle and a new platform for governing AI agents at scale.

The details:

• Cowork operates in the cloud, pulling from emails, meetings, files, and chats across M365 — a contrast to Claude Cowork's current desktop-only approach.

• Microsoft built Cowork directly with Anthropic, using Claude Cowork’s tech but wrapped in M365's enterprise security and compliance layers.

• Users describe an outcome, and Cowork breaks it into steps, producing deliverables like decks, briefing docs, and workbooks across apps.

• Cowork is available in a limited research preview, coming with a new $99/user E7 tier that bundles Copilot with agent management and security tools.

Why it matters: If you can't beat the thing that scared your investors, absorb it. Embedding Anthropic's agent tech inside M365's security boundaries gives Copilot Cowork something Claude Cowork can't easily match (yet) — deep, integrated enterprise context across 450M users’ worth of emails, calendars, and files.

AI TRAINING

📹 Auto-generate videos from content with Manus

The Rundown: In this guide, you will learn how to use Manus to script, generate, and assemble a short promotional video from any written content you already have — with no video editing experience required.

Step-by-step:

1. Start with a blog post, a press release, or even a news story. Download it as a PDF or Markdown file, and log in to Manus.im

2. Create a project with instructions: “You are a video producer for [my brand]. Your job is to transform written content into branded videos”

3. In the project task, click the plus button > Use Skills > video-generator, and prompt: “Create a 15–20 second teaser video based on the attached post. The CTA is to read the full blog. Generate it in portrait orientation for Instagram Reels. Give me four script options before you generate. Use default mode”

4. Carefully review Manus’s script options and pick the best one. Video generation should take 5–10 minutes and use between 499–800 credits

Pro tip: Save tokens by uploading B-roll videos and music into a Google Drive folder and connecting it to Manus.

PRESENTED BY GALILEO

⚙️ Expert guide to building RAG systems in 2026

The Rundown: RAG isn't dead, it's evolved: from basic vector search to today's agentic architectures with self-correction and adaptive retrieval.

Read Galileo's new 240+ page guide to learn how to build RAG systems that work in 2026, including:

• Mastering chunking, embedding, and reranking strategies

• Building evaluation frameworks using retrieval and generation metrics

• Implementing advanced patterns like query decomposition and adaptive retrieval

Get the free eBook.

AI RESEARCH

📊 a16z releases new consumer AI Top 100

Image source: a16z

The Rundown: a16z released the sixth edition of consumer AI Top 100, expanding the list to include traditional apps with AI like Canva and CapCut for the first time, along with data showing ChatGPT still dominates overall usage, but rivals are gaining ground.

The details:

• ChatGPT crossed 900M weekly users and still dwarfs every rival, but the gap is tightening — with Claude and Gemini growing paid subs over 200% last year.

• The new list included “AI-enhanced” consumer apps for the first time, with CapCut, Canva, Notion, Grammarly, and others now slotting into the rankings.

• The report found three distinct AI ecosystems forming: Western, Chinese, and Russian, with sanctions accelerating the split as local alternatives fill the gaps.

• Agents are gaining ground, with Manus (#44) and (#47) Genspark making the cut, while OpenClaw is absent due to the report’s time frame.

Why it matters: a16z's consumer reports have become one of the best pulse checks on where AI adoption is actually heading, and this edition is no different. Given the recent OAI Pentagon drama, cancellations, and Claude surge, the battlefield for consumers’ ‘default AI’ could be even more competitive in the next release.

QUICK HITS

🛠️ Trending AI Tools

• 🔒 Incogni - Remove your personal data from the web so scammers and identity thieves can’t access it. Use code RUNDOWN to get 55% off*

• 🤖 Copilot Cowork - Microsoft's Anthropic-powered AI for M365 tasks

• 🧠 GPT-5.4 - OpenAI's flagship reasoner with native computer use, 1M context

• 🗂️ Claude Marketplace - Anthropic's hub for Claude-powered partner tools

*Sponsored Listing

📰 Everything else in AI today

Anthropic rolled out Code Review for Claude Code in Team and Enterprise accounts, which uses teams of AI agents to deep-read code and flag bugs.

OpenAI announced the acquisition of Promptfoo, an AI security and red-teaming platform, to embed native agent testing into its Frontier enterprise platform.

Andrew Ng released Context Hub, a free tool that gives AI coding agents access to current documentation to prevent them from using outdated or hallucinated code.

OpenAI is further delaying its "adult mode" feature for ChatGPT, shelving the verified-users-only option to focus on intelligence, personality, and proactive capabilities.

Anthropic launched Claude Marketplace in limited preview, letting enterprises apply existing spend commitments toward partner tools from GitLab, Harvey, and others.

COMMUNITY

🤝 Community AI workflows

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 Tina J. in Farmingdale, NY:

"I’m working towards a promotion at the end of the year and used Copilot to evaluate my resume against real work artifacts and a competency framework. Copilot analyzed my work output and identified where I was already operating at the next level and where I could strengthen my visibility and framing.

Then I asked it to generate promotion-aligned custom instructions to reinforce strategic, executive-level behavior in future work."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: OAI robotics lead exits over Pentagon deal

• Read our last Tech newsletter: Meta sued over Ray-Ban privacy

• Read our last Robotics newsletter: DJI pays $30k for robot vacuum hack

• Today’s AI tool guide: Auto-generate video from writing with Manus

• RSVP to our next workshop Thursday @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, robotics enthusiasts. DJI is paying a French hobbyist $30K after he accidentally hijacked thousands of Romo robot vacuums — and the live camera feeds and microphones that came with them.

A single cloud permission flaw turned a wacky PlayStation controller experiment into an accidental mass surveillance incident. Patches are out, but the bigger question lingers: can we ever make smart devices truly private?

In today’s robotics rundown:

• DJI pays $30K for robot vacuum hack

• Alphabet’s delivery drones now work nights

• Ex-Googler launches robotics startup in Tokyo

• Tesla’s Full Self-Driving hits 8B miles

• Quick hits on other robotics news

LATEST DEVELOPMENTS

DJI

💵 DJI pays $30K for robot vacuum hack

Image source: Ideogram / The Rundown

The Rundown: DJI is reportedly paying $30K to the French hobbyist who accidentally seized control of 7K Romo robot vacuums, flagging a bug that turned an AI-assisted weekend hack into a global home‑surveillance scare.

The details:

• Sammy Azdoufal accidentally gained remote access to 7K robot vacuums worldwide while trying to control his own unit with a PlayStation controller.

• The flaw let him tap into live camera feeds, microphones, and detailed home floor maps from strangers’ devices via DJI’s cloud infrastructure.

• After he reported the issue and it went public, DJI pushed rapid backend fixes and automatic updates, saying the main vulnerabilities are now patched.

• DJI has agreed to pay Azdoufal $30K, positioning it as a bug-bounty style reward and vowing more third-party security audits and certifications for Romo.

Why it matters: This story is a test of how much trust we put in AI-powered gadgets that can see, hear, and map our homes, in the same line as Roomba test vacuums leaking intimate photos and Ring cameras letting strangers talk to kids in their bedrooms. It’s more proof that one bad cloud permission can go terribly wrong at scale.

WING

🥡 Alphabet’s delivery drones now work nights

Image source: Wing

The Rundown: Alphabet’s Wing just won the FAA’s approval to extend its drone delivery hours to 9 p.m. in parts of Dallas–Fort Worth and Charlotte, pushing on-demand aerial logistics firmly into the evening routine.

The details:

• Alphabet’s Wing now has FAA approval to run drone deliveries after sunset, extending operations into nighttime hours.

• Wing’s service for Walmart and DoorDash customers runs from 9 a.m. to 9 p.m. at select sites in the Dallas–Fort Worth metroplex and the Charlotte region.

• The drones use near-infrared “headlights” to see in the dark without adding visible light pollution, and can still pick safe landing spots, avoiding obstacles.

• Rival Zipline has now crossed 2M commercial deliveries, while Wing has logged roughly 350K drops and aims to cover 270 Walmart stores by 2027.

Why it matters: Amazon and Zipline are also chasing drone delivery at scale, but approvals for fully automated, after-dark flights are still rare enough that each new waiver quietly resets the bar for everyone else. Knowing when Wing slips into the night shift is also a useful tell for how fast regulators are loosening the reins.

INTEGRAL AI

🇯🇵 Ex-Googler launches robotics startup in Tokyo

Image source: Integral AI

The Rundown: Ex-Google researcher Jad Tarifi is turning Integral AI from a split SF–Tokyo business into a Tokyo‑first startup, betting Japan’s factory‑robot giants are the fastest way to put his AI “brains” into the real world.

The details:

• Tarifi has built a 15-person startup in Tokyo to plug advanced AI models into Japan’s huge industrial robot ecosystem.

• Integral AI has worked with auto parts giant Denso since 2021, using imitation learning so factory robots can pick up new tasks by watching human demos.

• The company is in talks with Toyota, Sony, Honda, and Nissan to show how language prompts let robots teach themselves complex workflows on the fly.

• Integral has raised about $5.5M so far and is seeking roughly $10M more to scale its models and launch its Genesis system later this year.

Why it matters: Japan controls a huge slice of the world’s industrial robots but still depends on foreign AI and cloud providers to run them, which is exactly the gap Integral AI is trying to fill. The company’s goal is to provide the “Silicon Valley brains” that sit on top of Japanese hardware to power next-gen factory automation.

TESLA

🚘 Tesla’s Full Self-Driving hits 8B miles

Image source: Tesla

The Rundown: Tesla’s Full Self-Driving (FSD) crossed 8B cumulative miles this year, a staggering data haul that inches Elon Musk’s long-promised robotaxi future closer to technical plausibility even as it remains stuck in regulatory and safety limbo.

The details:

• Tesla is pursuing autonomy using a camera-only perception stack, in contrast to Waymo’s lidar sensors and detailed 3D mapping.

• FSD has evolved into a supervised driver-assist suite that can change lanes, navigate routes, steer, park, and drive itself across parking lots to the owner.

• The company’s expanding fleet, software rollouts, and periodic free FSD trials have accelerated usage from just 6M miles five years ago to billions today.

• Musk has said about 10B miles of training data are needed to justify large-scale autonomous deployment, a threshold Tesla could hit within the year.

Why it matters: Tesla’s 8B FSD miles are huge, but those miles still come from a supervised Level 2 system rather than a true robotaxi network. That leaves Tesla with a massive data advantage over rivals in training volume, while Waymo retains the edge in fully driverless deployment, regulatory approval, and real-world commercial autonomy.

QUICK HITS

📰 Everything else in robotics today

OpenAI’s robotics chief, Caitlin Kalinowski, resigned over what she says is a rushed, poorly safeguarded Pentagon deal to put the company’s AI on military systems.

A London surgeon remotely controlled a Toumai robot to remove a prostate 2,400 km away with only 48 ms latency, showcasing long-distance telesurgery’s viability.

The U.S. National Highway Traffic Safety Administration will host an autonomous vehicle safety forum on Tuesday with the CEOs of Waymo, Zoox, and Aurora.

Alphabet CEO Sundar Pichai’s new equity package, worth up to $692M over three years, is reportedly tied to boosting the valuations of Waymo and Wing.

Mirai Robotics raised a €4.2M ($4.9M) pre-seed round to develop autonomous maritime systems aimed at both civilian and defense applications.

Agility Robotics is dropping “Robotics” from its name to become Agility, saying the rebrand reflects its readiness to scale humanoid deployments and lead adoption.

Ukraine is rolling out remote-controlled and autonomous armed ground robots against Russian forces, raising fresh ethical and legal concerns about AI-driven warfare.

Dexterity introduced Foresight, a physics-consistent world model that powers its Mech robot to autonomously load trucks using a 4D box-packing agent.

Ouster said demand for its lidar-powered software is taking off, with 2025 orders doubling and its systems now booked for more than 1,200 sites worldwide.

China’s new 5-year plan doubles down on rare earths and advanced robotics to secure supply chains and keep its edge as a high-tech industrial powerhouse.

NASA’s 1.8-meter Valkyrie humanoid is heading back to Johnson Space Center after a decade at the University of Edinburgh.

China’s WeRide is deepening its partnership with Geely’s Farizon unit to mass-produce 2K upgraded GXR robotaxis in 2026.

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: OpenAI's robotics lead exits over DoW deal

• Read our last Tech newsletter: Meta sued over Ray-Ban privacy

• Read our last Robotics newsletter: Waymo has a big problem in Austin

• Today’s AI tool guide: Build an AI case study generator

• RSVP to our next workshop Thursday @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team

Good morning, AI enthusiasts. The consumer backlash over OpenAI's Pentagon deal was loud, but this weekend marked an internal escalation: the company's robotics hardware lead resigned on principle.

Caitlin Kalinowski's departure is the first senior-level exit tied directly to the deal, and her public statement citing concerns over surveillance and lethal autonomy hits a bit harder than any App Store chart.

In today’s AI rundown:

• OAI's robotics lead exits over Pentagon deal

• The Rundown Roundtable: Our AI use cases

• How to build an AI case study generator

• Claude digs up 22 Firefox security flaws in two weeks

• 4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

OPENAI

🚪 OAI's robotics lead exits over Pentagon deal

Image source: Lovart / @kalinowski007 on X

The Rundown: OpenAI robotics director Caitlin Kalinowski just announced her resignation over the company’s controversial Pentagon deal, calling it a rushed move that skipped the guardrails on AI surveillance and lethal autonomy.

The details:

• Kalinowski joined OAI from Meta's AR glasses team in November, spearheading the rebuild of its robotics division that had previously shut down in 2020.

• She called the decision "about principle, not people", saying the deal was pushed through "without the guardrails defined" on AI in warfare.

• Kalinowski marks the first public “resignation” over the Pentagon deal, though VP of Research Max Schwarzer also departed last week for Anthropic.

• The backlash has hit fast on the consumer end, with Claude climbing to No. 1 on the App Store and ChatGPT cancellations soaring.

Why it matters: Plenty of users have ditched ChatGPT and spoken out since the Pentagon deal dropped — but Kalinowski is the first senior OAI voice to walk over it on principle. OAI can weather angry tweets and App Store slides, but a big resignation letter that name-drops "lethal autonomy" and "surveillance" hits a bit differently.

TOGETHER WITH YOU.COM

📖 AI Agents are your new employees

The Rundown AI agents are entering the workforce, but you wouldn’t expect a new employee to know everything on day one, would you? AI agents need onboarding too—in the form of metadata.

In this eBook, you’ll learn:

• How metadata management drives AI success

• Common pitfalls

• The ROI of proper metadata management

Get the eBook.

THE RUNDOWN ROUNDTABLE

💡The Rundown Roundtable: Our AI use cases

The Rundown: The Rundown Roundtable is a weekly feature where we poll members of The Rundown staff about how we use AI in our work and daily lives.

Rishi, Growth: We’re always looking for A/B test ideas for our landing pages and new ways to improve conversion rates while creating a better user experience. Recently, I tried an interesting AI workflow after coming across a public CRO skill on Twitter.

I installed it in Claude Code and then fed Claude screenshots of our landing page along with behavioral data from Microsoft Clarity, including scroll depth, heatmaps, and which buttons people were clicking most.

Using the CRO scorecard framework and the Clarity data, Claude generated a detailed analysis of the page and recommended five A/B tests we should prioritize, along with the reasoning behind each one. The insights were genuinely useful, and we have already taken action by launching one of the recommended tests.

Adrian, Developer: I’ve basically had the exact same haircut since high school, so this year I finally decided to experiment a bit with Nano Banana 2 and see what a different look might feel like. I uploaded my own portraits and started merging them with different models’ hairstyles. After 10+ rounds of virtual makeovers, I found out that the hairstyle that suited me best was… my current one.

AI TRAINING

⚙️ Build an AI case study generator

The Rundown: In this guide, you will learn how to use Claude to turn those old project files, like client emails, metrics exports, or memos, into information-rich case studies you can use to win more proposals.

Step-by-step:

1. Write project wrap memos when a project ends. Give an overview of results and key KPIs. If you’re busy, even a quick voice memo is better than nothing

2. Create a “Case Study Generator” project in Claude’s PC app with instructions: "You are a case study writer for [industry]. Your job is to take raw data and turn it into a case study using the challenge → solution → results framework"

3. Upload your project memo and prompt: “Generate a case study from these docs. Lead with strongest results in the headline. Use the challenge → solution → results framework. If any section is missing data, flag it with [NEEDS INPUT]”

4. Claude should generate a PDF case study for you. We found that it did a great job turning the memo into a slide deck or social media carousel post

Pro tip: We can pack a lot more into our system instructions, including a style guide and example cases. You can grab the templates here.

PRESENTED BY IBM

💪 How to build a strong AI-ready data foundation

The Rundown: IBM explains how unified, secured, and governed data access can help organizations move promising AI pilots to reliable enterprise scale.

In this guide, you’ll:

• Understand barriers that prevent AI pilots from scaling

• Learn why unified access to structured and unstructured data matters

• Explore a framework for building AI-ready data foundations

Download the guide.

ANTHROPIC & FIREFOX

🔒 Claude digs up 22 Firefox security flaws in two weeks

Image source: Anthropic

The Rundown: Anthropic revealed that Claude Opus 4.6 spent two weeks tearing through Firefox's codebase alongside Mozilla's team, turning up 22 vulnerabilities (14 high-severity) — with patches already live for hundreds of millions of users.

The details:

• Claude took just 20 minutes to flag its first flaw, and racked up 50 more by the time Anthropic's team finished confirming its initial find was real.

• Anthropic filed 112 reports across ~6K files in total — 14 rated high-severity by Mozilla, accounting for nearly 20% of Firefox's most serious patches all year.

• Claude also tried writing exploits, but only managed two working attacks in hundreds of attempts — both needing Firefox’s sandbox removed to function.

Why it matters: Firefox isn’t some new app; it’s a deeply tested open-source project with decades of audits and bounty programs — making Claude’s quick findings even wilder. While Claude wasn’t as strong at weaponizing its own exploits, Anthropic said that gap won’t last… Meaning the window to lock down codebases feels pretty urgent.

QUICK HITS

🛠️ Trending AI Tools

• 🗣️ Unwrap Customer Intelligence - Connect your entire organization to the true voice of the customer with AI-driven insights from customer feedback*

• 🔒 Codex Security - OpenAI's security agent to scan repos and patch bugs

• ⚙️ autoresearch - Andrej Karpathy's tool for AI-driven LLM training

• 🎆 Uni-1 - Luma’s model that reasons and generates across text and images

*Sponsored Listing

📰 Everything else in AI today

Luma unveiled Uni-1, its first model that combines reasoning and image generation in one architecture — in a major shift from the video-focused startup's roots.

Anthropic rolled out scheduled tasks in Claude Code, letting the coding agent run prompts on a loop to monitor builds, check logs, and auto-file PRs on a set cadence.

Cluely CEO Roy Lee admitted to fabricating the startup's revenue figures in a 2025 interview, publicly retracting claims after the AI ‘cheating’ tool pivoted to meeting notes.

The WSJ shared more on AI's role in the Iran conflict, reporting that the Army's 18th Airborne matched its Iraq-era targeting with 20 people instead of 2,000 with the tech.

Andrej Karpathy released autoresearch, an open-source repo that lets AI agents autonomously run and iterate on LLM training experiments in a loop on a single GPU.

COMMUNITY

🤝 Community AI workflows

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 Julie H. in Kyle, TX:

"I made a day planner agent connected to my calendar and task lists. Every morning, I start by asking, “How does my day look?”, and the agent pulls all my meetings, uses my tasks list to schedule projects in between meetings, gives me time for deep focus and email catch-up, and even makes sure to schedule a lunch break.

It flags things I may have missed the day before. If I have big gaps in the day, it looks ahead and suggests items I can get ahead on, like tasks due the next day or meetings needing prep time. I can give feedback on things I want to add, move, or remove, and it adjusts until I have a solid plan for my day. If I move things in my schedule, it will also update my calendar and task list for me so everything stays aligned."

How do you use AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: OpenAI’s ‘best model ever’ goes live

• Read our last Tech newsletter: Meta sued over Ray-Ban privacy

• Read our last Robotics newsletter: Waymo has a big problem in Austin

• Today’s AI tool guide: Build an AI case study generator

• RSVP to our next workshop Thursday @ 2PM EST: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, AI enthusiasts. A week ago, Spotify’s co-CEO claimed their best devs haven’t written a “single line of code” this year — echoing a wave of execs that describe AI coding agents as the future of software. 

The shift is happening – there’s no doubt about that. But bringing AI into real engineering workflows is more nuanced than hitting a switch and going on autopilot.

To better understand what’s changing (and what isn’t), we sat down with Rajeev Rajan, CTO of Atlassian, the company behind popular collaboration tools such as Jira, Confluence, and Loom. We got insights on their recent software development agent Rovo Dev and discussed how human roles evolve in an AI-native world.

In today’s AI rundown:

• The non-negotiable cost of coding with AI

• When AI writes code, what will an engineer do?

• Building agentic AI for developer joy

• What happens when AI makes a mistake?

• Truth about the “death of SaaS” theory

• Quick hits with Rajeev

LATEST DEVELOPMENTS

WORKFLOW REDESIGN

🤖 The non-negotiable cost of coding with AI

The Rundown: As more enterprises adopt coding agents, Rajan says teams will have to redesign their workflows with both human ownership and safety systems, so that AI moves faster without sacrificing consistent, production-grade quality.

Cheung: AI is generating more code than ever, but research shows 45% of it still contains security flaws. How exactly should teams leverage AI coding agents without trading quality for speed?

Rajan: There’s an undeniable trend toward an AI‑native software development lifecycle. But if you let quality slide, you’re just moving faster toward incidents and customer pain. It’s less about “AI vs. quality” and more about “how do we redesign the workflow on the assumption that AI is in the loop by default?”

In code review, we are entrusting AI to catch bugs, enforce coding standards, and explain complex changes. For example, Rovo Dev helped reduce PR cycle time by 45% and auto-resolved 51% of potential security vulnerabilities. The nature of review is changing here: instead of humans reading every line of a peer’s code, it’s about a human owner reviewing an agent’s work.

Rajan added: At deployment, if AI is helping you generate and ship more code, your safety systems have to keep pace. Think: smaller batches, heavier CI, stronger observability, and fast rollbacks. You can’t operate in a black‑box scenario.

Why it matters: Speed only becomes an advantage when the system behind it is airtight. The right approach with AI coding agents is to treat them as core enterprise infrastructure — which means designing workflows around them, building in safety from day one, and holding output to the same standards as human-written code.

ROLE SHIFT

❓️ When AI writes the code, what will an engineer do?

Image: Kiki Wu / The Rundown

The Rundown: With writing code no longer the bottleneck, Rajan believes that the next big opportunity for engineers is stepping into more strategic functions (like planning to execution) and designing better systems with AI in the loop.

Cheung: How much of code will be AI-written by 2028, and what does the role of a software engineer look like with that change?

Rajan: By 2028, I would not be surprised if most new code in large companies is AI-generated. I say that as someone who fell in love with writing code early in my career and still remembers the joy of seeing something work for the first time.

We’re seeing a shift where every engineer is a tech lead, orchestrating systems and agents. Engineers now spend more time driving clarity and owning what happens “left of code” and “right of code” – from planning and design on one side to testing, rollout safety, and operations on the other.

Cheung: What about new grads entering the field — does AI help them or hurt them?

Rajan: Focusing on the right fundamentals and adopting the AI-native way of working will give new grads a big advantage — potentially allowing them to leapfrog senior developers who haven’t adopted AI ways of working yet. Your edge will come from judgment: knowing when to trust the AI and when to challenge it.

Why it matters: As AI writes more code, the moat for engineers moves from actually typing to instead framing problems, designing systems, and maintaining oversight. Rajan’s leapfrog point is especially interesting: the advantage may not go to the most senior person, but to whoever learns to orchestrate AI fastest.

DEVELOPER JOY

⚙️ Building agentic AI for developer joy

The Rundown: Atlassian kicked off its internal journey to improve “Developer Joy,” raising developer satisfaction scores from 49% to 83%. With teams moving faster and feeling more empowered to make changes, Rajan shared how this renewed sense of ownership led to direct product improvements with Rovo Dev.

Cheung: Why did you decide to focus on developer joy, and how did you actually measure it?

Rajan: When I joined Atlassian, we chose to frame developer productivity as ‘Developer Joy’. If developers are frustrated, blocked, or taken out of their flow, it doesn’t matter what productivity metric you pick — you’re not going to get great outcomes.

We track this with regular satisfaction surveys and hard metrics tied to pain points. Developer satisfaction has gone from 49% to 83%, and we see that show up in the work. For example, by focusing on one of the biggest friction points, the Confluence backend team cut its full server build time by more than 60%. We see these kinds of investments as core to our ability to ship value to customers.

Cheung: When you tested early versions of Rovo Dev internally, what did engineers push back on?

Rajan: Early on, the feedback we got on Rovo Dev was that parts of the experience felt like ‘magic’ in the wrong way. It would do something useful, but you could not see enough of the work it did to get there.

We actually scrapped and reworked an early ‘one click, do it all’ flow because our own teams would not touch it without more transparency and control. They wanted a way to understand each step the agent took, how their instructions led to different outcomes, and the ability to steer the agent. That pushed us hard toward agent sessions you can inspect and experiences that keep developers in the loop.

Why it matters: Rajan’s early Rovo Dev story highlights how critical internal feedback loops are when adopting agentic AI. The more teams listen to users (and iterate on what feels opaque, risky, or frustrating), the stronger and more trustworthy the system becomes. Iteration goes hand-in-hand with developer trust in an AI-native world.

ACCOUNTABILITY

🧠 What happens when AI makes a mistake?

The Rundown: Rajan says powerful AI agents should be deployed to production only when there’s clear human ownership and a way to track, monitor, and steer the system’s behavior, creating an accountability layer that moves as fast as the AI itself.

Cheung: As AI agents take on more autonomous work, how do you maintain accountability, and why can’t “the AI did it” ever be an excuse?

Rajan: When something goes wrong, ‘the AI did it’ can’t be the answer, because AI doesn’t own customer trust – we do.

As we bring autonomous AI into our workflows, we have to be explicit about accountability: every AI-assisted decision or action has to have a clear human owner. If we can’t understand or observe how an AI is behaving, it doesn’t belong in a critical path.

We put guardrails and observability around AI, we log and audit its actions, and we make sure teams treat it like any other powerful tool: you understand failure modes, you monitor it, and you don’t ship without ownership. AI can help move faster, but it doesn’t replace judgment and responsibility.

Why it matters: With something as powerful as agentic AI, accountability is non-negotiable. If you get it right, you unlock speed, trust, and durable customer confidence. But if you get it wrong, the consequences can be massive — because autonomous systems can amplify mistakes just as quickly as they amplify progress.

SAAS STORY

⚡️ Truth about the “death of SaaS” theory

The Rundown: Despite the rise of AI coding agents, Rajan argues SaaS tools aren’t going anywhere. In fact, he believes these tools will get stronger — with AI working across projects and controls by tapping contextual insights.

Cheung: What’s your take on the whole “saaspocalypse” narrative that AI agents will kill SaaS altogether?

Rajan: The idea that one person in your team can vibe code an in-house solution over a weekend and replace a mature SaaS solution you’re paying for, in my view, is overrated.

When customers buy software-as-a-service, they are not just buying the code; they buy workflows, shared context, security, compliance, and reliability. That is where well-designed SaaS products still matter a lot. What AI actually does is make great SaaS even more valuable.

Rajan added: Your projects, docs, tickets, and conversations live in those systems, and AI can now move across them, automate the boring parts, and orchestrate agents around trusted workflows and controls. So I am more interested in SaaS that becomes AI-native than in hot takes that SaaS is dead.

Why it matters: The debate is still on, but one thing’s clear: as AI agents grow more capable, the foundation of SaaS will become even more important. AI will evolve the platforms that already hold an organization’s workflows and institutional knowledge, serving as trusted systems of record.

LIGHTNING ROUND

⚡️ Quick hits with Rajeev

Most underrated AI trend right now?

Rajan: Lots of AI products today are designed for a single-player system. We see much greater potential in how AI helps entire teams work better together — allowing important context to flow across a team of humans and agents.

Something you believe about AI that most people in tech would disagree with?

Rajan: I think AI will make engineering more human, not less. A lot of people worry we will lose the craft — I believe we will spend less energy on repetitive implementation and more time on strategic, creative work and collaboration.

Advice for teams struggling with developer burnout?

Rajan: Start by fixing one concrete, high-friction problem that impacts your team. You’d be surprised by how quickly chipping away at issues like slow build times and noisy tooling can multiply and have a greater impact.

You worked at Microsoft for 20+ years, then led engineering at Meta. What did each teach you about building great teams?

Rajan: Microsoft taught me the value of deep technical rigor and building platforms that stand the test of time. Meta taught me how powerful it is when you pair strong engineering talent with a bias for fast iteration and learning. At Atlassian, I try to combine both: long-term architecture with a culture that ships, learns, and adapts quickly.

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown

Good morning, tech enthusiasts. Meta’s Ray-Ban smart glasses are at the center of a nightmarish privacy lawsuit: a proposed U.S. class action alleges the company let human reviewers access users’ intimate recordings, despite marketing that suggested stronger protections.

The case is raising a simple, unsettling question for smart-glasses users: when the camera is on, who might really be watching?

In today’s tech rundown:

• Meta’s AI glasses hit with privacy suit

• Apple’s $599 MacBook for Chromebook crowd

• Oura acquires gesture-recognition startup

• Science Corp nabs $230M for brain implant

• Quick hits on other tech news

LATEST DEVELOPMENTS

META

👓 Meta’s AI glasses hit with privacy suit

Image source: Ideogram / The Rundown

The Rundown: Meta is facing a proposed U.S. class-action lawsuit alleging that its Ray-Ban Meta AI smart glasses secretly routed intimate user videos to low-paid human reviewers, violating the company’s own privacy promises.

The details:

• A proposed U.S. class-action lawsuit accuses Meta and Luxottica of misleading buyers about the privacy protections of Ray-Ban smart glasses.

• Plaintiffs say Meta failed to disclose that contractors and employees could view user recordings, including nudity, sex acts, and people using the bathroom.

• The case follows reports that low-paid reviewers in Kenya were required to watch and label intimate clips as part of Meta’s safety and AI-training pipeline.

• Regulators, including authorities in the U.K., are also probing whether Meta’s data practices for the glasses meet transparency and consent standards.

Why it matters: The complaint says Meta misled customers about this human-review pipeline and put millions at risk of stalking, extortion, and identity theft, while Meta counters that contractors review shared Meta AI content to improve the glasses’ user experience and that this human involvement is disclosed in its policy fine print.

APPLE

🍏 Apple’s $599 MacBook for the Chromebook crowd

Image source: Apple

The Rundown: Apple launched a $599, iPhone-chip-powered MacBook Neo that brings bright colors and bargain pricing to directly challenge low-cost Windows laptops and Chromebooks. It’s set to hit stores on March 11.

The details:

• The Neo uses an iPhone‑class A‑series chip paired with up to 16GB of unified memory and fast NVMe storage, aiming for all‑day battery life.

• Apple is offering the Neo in multiple bright colors, visually positioning it closer to iPads and older plastic MacBooks than its current pro-focused lineup.

• The machine targets students and first-time buyers who currently gravitate toward Chromebooks and cheaper Windows laptops.

• By undercutting traditional MacBook prices, Apple is moving aggressively into a segment long dominated by PC OEMs and Google’s education partners.

Why it matters: Apple is putting real pressure on bargain Windows laptops and Chromebooks with a $599 MacBook that still feels like a “real” Mac. By pairing an iPhone-class chip with a colorful, fanless design and long battery life, the MacBook Neo aims to hook first-time buyers without undercutting its pricier line.

OURA

👊🏼 Oura acquires gesture-recognition startup

Image source: Oura

The Rundown: Smart ring maker Oura just acquired Helsinki-based gesture-recognition startup Doublepoint to fuse voice and micro‑hand movements into its rings, turning the health tracker into a discreet control hub for next‑gen wearable AI.

The details:

• The deal is designed to layer gesture recognition and voice controls on top of Oura’s continuous health sensing to trigger features without screens.

• All of Doublepoint’s team, including its four co-founders, will join Oura, forming a new hub for interaction design inside the Finnish wearable company.

• Oura, valued at $11B, has sold more than 5.5M rings and expects sales to top $1.5B in 2026 as smart ring shipments jumped about 51% last year.

• This is Oura’s fourth recent acquisition, following Sparta Science, Veri, and Proxy, each adding a new layer to what was primarily a sleep tracker.

Why it matters: Oura CEO Tom Hale said gesture controls, paired with voice inputs, could become central to future versions of the ring, though meaningful implementation isn’t happening right away. Still, the acquisition helps Oura defend its smart ring lead in the category, as wearables start moving toward a post-screen interaction model.

SCIENCE CORPORATION

🧠 Science Corp nabs $230M for brain implant

Image source: Science Corp (PRIMA retinal implant)

The Rundown: Neuralink co-founder Max Hodak’s neurotech startup Science Corporation closed a $230M Series C, valuing the company at roughly $1.5B and putting serious fuel behind its bid to restore sight to patients blinded by retinal disease.

The details:

• The funds will accelerate clinical trials and regulatory approvals in the U.S. and Europe for PRIMA, a subretinal chip paired with camera-equipped glasses.

• In early trials with 47 patients, PRIMA has helped people with advanced macular degeneration recognize letters, numbers, and short words.

• Science Corp. is seeking a CE mark in the EU, aiming for mid‑2026 approval with Germany as its first launch market, while it continues talks with the FDA.

Why it matters: Science Corporation’s side projects — from cortical interfaces to organ-preservation tech — show it’s thinking beyond implants to a neurotech platform. That puts it in more direct competition with Neuralink and Precision Neuroscience, which are also racing to own the hardware layer between brains and machines.

QUICK HITS

📰 Everything else in tech today

Oracle is reportedly preparing to cut thousands of jobs as it grapples with a cash crunch caused by massive spending on AI data centers and related infrastructure.

A global Amazon outage left tens of thousands of shoppers facing failed payments and checkout glitches, with Downdetector logging around 20K problem reports.

Online retailer Quince is reportedly in talks with investors to raise a new funding round that would more than double its valuation to above $10B.

Microsoft confirmed that its next-gen Xbox, codenamed Project Helix, is a high-performance console that will run both Xbox and PC games.

U.S. defense tech firm Anduril is raising a multibillion-dollar round that could double last year’s $30B valuation to about $60B. 

Researchers built a stable, ultra-bright 300-nanometer OLED pixel using a nano-antenna, enabling full HD displays small enough to fit on a grain of sand.

London-based smartphone maker Nothing launched new phones and headphones in bright colors to counter what its founder calls a “boring” tech market.

Whoop, the maker of connected fitness bands, plans to increase its workforce by about 75% in 2026 by hiring more than 600 people to fuel growth ahead of a likely IPO.

Cluely co-founder and CEO Roy Lee admitted on X that the $7M annual recurring revenue figure he gave TechCrunch last summer was a deliberate lie.

BYD unveiled its Blade Battery 2.0 pack, which can fully charge in just about 10 minutes but only when paired with its new 1.5-megawatt Flash Charging stations.

COMMUNITY

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: OpenAI's 'best model ever' goes live

• Read our last Tech newsletter: Apple’s budget iPhone goes AI first

• Read our last Robotics newsletter: Waymo has a big problem in Austin

• Today’s AI tool guide: Turn an investment memo into a polished slide deck

• RSVP to our next workshop on March 12: AI coding bootcamp pt. 1

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — The Rundown’s editorial team