October 28, 2026

Global AI Edition

The past week’s cyber headlines have had an end-of-the-world quality, with security flaws sending governments and companies scrambling to take websites offline. A useful reminder to check our own digital locks. (Axios)

Start with the phone in your hand. If you have an iPhone, open the Passwords app and tap Security. See how many saved passwords are flagged as reused, weak or exposed in known data leaks. The number may be sobering. Your dog’s name has had a long career in cybersecurity. It’s time to let it retire. Tap each flagged account and follow the steps to change its password. (Apple’s guide)

Whatever devices you use, start with your email and banking accounts. Give each account a unique (strong) password, let a password manager remember it, turn on two-factor authentication, and install those software updates you keep postponing. Be suspicious of urgent texts or emails asking you to log in—open the app or website yourself. (CISA’s checklist)

You can’t patch the entire internet before lunch. You can stop using the same password for your bank and your pizza delivery.

Claude's New Employee

Anthropic's Duet Joins the Personal Agent Game, Taking on Your Inbox, Calendar and To-Do List (TechCrunch)

Anthropic is pushing deeper into personal AI agents with Claude Duet, a new Opus 5.5-powered app designed to proactively manage work, complete tasks, and keep going on its own cloud computer even after you close the app.

Duet carries over the memory, skills, and connectors from your Claude account, letting it work across tools like Gmail, Google Calendar, Slack, QuickBooks, HubSpot, DocuSign, and Calendly while using its cloud browser for sites without connectors.

You can use it to start each morning with a personalized brief, clear out low-priority inbox clutter, draft replies, chase unpaid invoices, prep research before meetings, schedule appointments, or hand off larger projects like financial models, slide decks, dashboards, and research reports.

Unlike a chatbot that waits for every prompt, Duet can monitor connected accounts at intervals, flag what needs attention, follow up on stalled work, and keep its progress organized in a dedicated Tasks tab. It can also make approved purchases through Link or Shop Pay using one-time cards, but Anthropic says Duet must ask before sending messages, making payments, signing documents, or deleting anything.

There are still limits: users can’t watch or take over its browser, it may miss urgent items between account checks, some connectors have restrictions, and it can make mistakes that require users to review drafts and approval requests carefully.

Claude Duet is currently rolling out by waitlist as a US-only research preview for Claude Pro and Max subscribers, giving early users a glimpse at what an AI assistant looks like when it moves from answering questions to actively handling the work around them.

Inside Enterprise AI

Industry Take: What the “SaaS is Dead” Narrative Misses (The Rundown AI)

The Rundown AI's founder Rowan Cheung sat down with CTO Philipp Herzig to talk about the “SaaS is dead” debate, how AI is changing the way people use work software, what agents need to know to get answers right, how to keep AI bills in check, and the models built for business data.

SaaS, or software-as-a-service, refers to cloud-based business software, and the “SaaS is dead” argument is that AI could make many traditional software products obsolete by letting people build their own tools on demand.
‌

As AI coding tools let almost anyone build apps, many believe traditional software companies are on the way out. Herzig thinks that narrative confuses building an app with actually running one — and says the same gap helps explain why so many AI pilots never make it into production.

Here's a conversation excerpt from the interview:

Cheung: What’s one AI trend you think is overhyped?

Herzig: The idea that “SaaS is dead” because AI can now write software. It confuses the ability to generate an application with the responsibility of running it.

Generating an expense app is one thing. Running it across 40 countries, keeping up with tax rules, enforcing permissions, and reconciling reimbursements is another. As agents take on more autonomous work, they depend more on trusted data, business context, governance, and security. The interface can change radically. The need for reliable business processes doesn’t.

Cheung: Why do so many AI pilots never get deployed?

Herzig: The difference between a prototype and a production-ready system at enterprise scale, with all the technological dependencies and complexity under the hood while still looking simple for the user, is still substantial human work.

Cheung: What’s been harder at SAP, the technology or adoption?

Herzig: Adoption, without a doubt. Useful AI changes who makes which decisions and how. That affects expertise, authority, and professional identity.

The companies moving fastest gave access to people whose judgment the organization already trusted. A respected supply chain manager saying “this helped me catch a disruption” is more powerful than any training program.

Why it matters: SAP has an obvious stake in defending enterprise software, but the gap between generating an app and running a business on it is real. If you’re rolling out AI at work, Herzig’s adoption advice applies regardless of vendor: pick one high-value use case and put the colleagues people already trust first in line.

Read more takes from Herzig in the full interview here.

What I'm Reading

IPO Vital Signs: Anthropic IPO Filing Calms Nerves With Signs of Healthy Growth (Bloomberg)

Anthropic’s IPO filing gave battered AI investors something to cheer about, revealing $16.3 billion in first-half revenue—up more than 1,300% from a year earlier—along with a broad enterprise customer base and $131.7 billion in cash.

The numbers helped revive enthusiasm across AI-related stocks, with investors particularly encouraged that Anthropic’s revenue is not overly dependent on a handful of customers and that its huge cash cushion could help fund future growth.

Financial Tech News: Bloomberg Introduces BQuant Assistant to Help Financial Professionals Accelerate Data Analysis and Coding Workflows (Bloomberg - Press Release)

Bloomberg today introduced BQuant Assistant, a domain-grounded conversational AI coding assistant built into BQuant Desktop and BQuant Enterprise.

“Financial professionals are working with increasingly large and complex datasets, but turning an investment idea into a repeatable analysis can still require significant time and specialized coding knowledge,” said Shawn Edwards, Bloomberg’s Chief Technology Officer.

Integrated within the BQuant development environment, BQuant Assistant gives financial professionals who are Bloomberg Terminal subscribers the ability to use natural-language prompts to write code, build dynamic applications, and accelerate financial data analysis workflows.

China's Frontier Push: Moonshot Debuts First AI Upgrade Since Its Own DeepSeek Moment (Bloomberg)

Moonshot AI, the Chinese upstart that had its own DeepSeek moment over the summer, released the first major upgrade to the seminal model that thrust the $50 billion company into the frontier.

The Beijing-based startup unveiled Kimi K3.1 Pro, the first in a new series of AI models that is designed to be significantly better at financial research and coding, two areas of focus for rival developers in the US.

The new release is intended to keep Moonshot competitive in shipping near-frontier AI models alongside competitors like DeepSeek and Z.AI Co. Upon its debut in July, K3 delivered performance on industry benchmarks that rivaled top-tier offerings from OpenAI and Anthropic. The surprise breakthrough — much like DeepSeek’s breakout R1 hit from early 2025 — sent AI and semiconductor stocks tumbling as investors feared it could make US data and computing buildouts obsolete.

Siri Learns New Tricks: Apple Rolls Out iOS 27.2 with New Siri AI Languages and Security Fixes (TechCrunch)

Apple’s iOS 27.2 is a wide-ranging upgrade that makes Siri AI more useful and more personal, expanding the assistant to French, Japanese, Korean, Portuguese, and Spanish in beta while giving it the ability to remember past conversations and pull relevant details from Messages, Calendar, voicemail, Safari, and other sources.

The update also brings Apple Pencil support to the foldable iPhone Duo, adds FaceTime features such as simultaneous front-and-rear camera streaming and support for third-party sign language interpreters or live captioners, and introduces smarter message suggestions, Podcast updates, and Multiview for watching up to four live sports streams at once.

Security is a major part of the release as well, with more than 160 fixes included in iOS 27.2 and many of them also backported to iOS 26, which Apple says is still running on about 80% of iPhones.

Choose Your Own Reality: Google Upgrades AI World Model That Rattled Video Game Stocks (Bloomberg)

Alphabet's Google announced a new version of its artificial intelligence world model that generates more realistic, real-time virtual environments, advancing a technology that has rattled investors in video game companies.

Genie 4, unveiled Tuesday by the company’s Google DeepMind AI lab, lets users create explorable worlds from text prompts and navigate them with keyboard controls. People can also alter a scene as it unfolds by typing instructions, such as changing the weather or adding a character.

Design Intelligence: Canva Releases Its First Image Generation Model (TechCrunch)

Canva is pushing deeper into generative AI with Canva Craft, its first homegrown image-generation model, giving the platform a new way to help users create everything from social posts and ads to presentation visuals, posters, product shots, and editorial-style images.

Built specifically for design, Craft can generate high-quality images up to 2K resolution in more than 30 shapes and sizes while doing a better job with details like readable text, object placement, and overall layout.

Because it was trained heavily on millions of human-made Canva templates rather than mainly on photographs, its images can also be turned into editable design elements with tools like Magic Layers, making it easier to tweak one part of a graphic without rebuilding the whole thing.

At launch, Craft debuted fourth among individual image-generation models on DesignArena, making Canva Research the platform’s second-highest-ranked image-generation lab.

You can try Craft now in Canva’s Dream Lab, where a simple prompt can be turned into anything from a polished social graphic to a product image, poster, or presentation visual.

Robot Boot Camp: In the A.I. ‘Robogym,’ Machines Train to Be More Like Humans (New York Times)

A little-known German robotics company is suddenly attracting some very big attention: Neura Robotics has raised as much as $1.4 billion and drawn backing from Nvidia, Amazon and other major investors as Europe races to put AI to work in the physical world.

Inside Neura’s 24/7 “robogyms,” humanoid and industrial robots practice factory skills like moving car parts, stacking boxes and learning how tightly to grip objects—then upload what they learn to a cloud-based system so other robots can get smarter, too.

The race is lopsided: China already operates more than 90 humanoid data-collection centers and has deployed 2.3 million industrial robots, while Europe has fewer than 10 comparable training facilities. Neura plans to have five robot gyms operating globally by year-end and is also equipping more than 1,000 workers with sensor suits to capture real-world movement data that can help train the machines.

Founder David Reger believes this new generation of robots could help revive Europe’s manufacturing muscle, but first they’ll have to prove they can do the job safely, efficiently and cheaply enough to compete at scale.

Photo: Roderick Aichinger (NYT)

Best of Late Night

On former President Barack Obama mocking Trump’s White House renovations, saying,“He should be on HGTV. He’s got the White House looking like Saddam Hussein opened a Cheesecake Factory. Just gold everywhere":

“So then Fox News, of course, jumped on that. They called Obama an ‘out-of-touch elitist’ for making fun of the Cheesecake Factory. Right, but the guy building the 90,000-square-foot ballroom, man of the people, absolute salt of the earth.” 
-- Jimmy Kimmel

“The part of the comparison that bothered you was the Cheesecake Factory part? Not the Saddam Hussein? You know, the Cheesecake Factory is not an actual factory, right?"
-- Jon Stewart

“I think there is only one way to settle this: roast battle on the White House lawn, live. I’ll call Jeff Ross and make this happen tomorrow.”
-- Jimmy Kimmel

And from The Daily Show: With the 2026 midterms closing in, Republicans are facing an unpopularity crisis as gas prices skyrocket and the war in Iran drags on. In this new special, Jordan Klepper hits the MAGA campaign trail to see if anti-communist hysteria, God's pronouns, and old-school Islamophobia can hold the GOP's fraying coalition together. Watch the clip by clicking the image below. 

Thank you for reading!

Please forward my email to anyone you think should read it. They can sign up here. If you no longer wish to receive the newsletter, you can unsubscribe here.

Follow me on X and LinkedIn for more updates. And get the latest updates from Mike on LinkedIn, X, Instagram and Threads.

Reply to this email with any feedback or tips, or send a note to dailyread@bloomberg.net.