In partnership with
In Today’s Issue:
🏛️ Sam Altman defends OpenAI’s aggressive burn rate in a new Forbes interview
🚀 Altman appears at the White House alongside Masayoshi Son to announce a $500 billion joint venture to build 20 massive AI data centers across the U.S
🧬 GPT-5.2 achieves a 40% reduction in protein synthesis costs
🦞 As OpenAI rolls out a budget-friendly $8/month tier with sponsored content
🤖 In a bold philosophical move, Altman declares he would "gladly step aside" if an AI system becomes capable of running OpenAI as CEO
✨ And more AI goodness…
Dear Readers,
The $285 billion question isn't whether AI will disrupt software, it's whether anyone saw it coming this fast. When Anthropic released legal automation plugins last week, Wall Street didn't just blink; it hemorrhaged a quarter-trillion dollars in market value as investors suddenly realized foundation models aren't just enhancing software anymore, they're replacing it. Thomson Reuters crashed 18%, LegalZoom fell 20%, and even Microsoft took a beating - all because AI agents can now do the grunt work that justified thousand-dollar-per-seat subscriptions.
Today we're digging into the shakeup that has SaaS executives sweating, exploring how deepfake fraud just hit industrial scale with a 700% surge, and examining why AI mathematicians are cracking problems humans couldn't solve in decades. Plus: Claude's new Fast Mode, the trust crisis brewing in every video call you take, and why three out of ten fraud attempts now use AI-generated content. The platform layer is eating the application layer in real time; let's break down what that means for everyone building, investing, or working in this space.
All the best,
🚨 AI Legal Tool Shakes Markets
A new legal automation plugin from Anthropic triggered a sharp market selloff, highlighting how fast AI can spook investors worried about mass disruption across law, finance, and SaaS. Software stocks slid hard - Thomson Reuters plunged over 20%, while Salesforce and CrowdStrike also dropped—despite analysts warning the fears may be overblown given AI’s real-world limits. Still, as Reuters reports, even incremental AI upgrades are now powerful enough to move billions in market value overnight (more on that in todays daily Topic down below)
🤖 AI Solves Longstanding Math Mysteries
A new AI startup called Axiom claims its system has solved four previously unsolved math problems, including a thorny conjecture in algebraic geometry that human mathematicians couldn’t fully crack. The breakthrough suggests AI reasoning is moving beyond pattern-matching into genuine mathematical insight, potentially accelerating discoveries that once took years - or decades - of human effort. If verified, this could reshape how advanced research in math and theoretical science gets done.
⚡ Faster Claude Responses With Fast Mode
Fast mode is a research-preview setting that speeds up Opus 4.6 responses by prioritizing low latency over cost efficiency, making it ideal for rapid iteration, live debugging, and time-sensitive work. It delivers identical quality to standard mode but charges higher per-token rates, with a limited-time 50% discount available until February 16. You can toggle it on anytime in Claude Code, though enabling it early in a session is the most cost-efficient move. However, its very pricy - so be careful!
AI's Research Frontier: Memory, World Models, & Planning — With Joelle Pineau
The Democratization of Deception: AI Scams Hit Industrial Scale
The Takeaway
👉 Deepfake fraud has industrialized: AI tools now enable anyone to create convincing fake videos, voices, and identities at minimal cost, with deepfake fraud surging 700% in Q1 2025 and three in ten retail fraud attempts now AI-generated.
👉 Traditional detection methods are failing: Scammers eliminated spelling errors and awkward phrasing through AI, making phishing emails and voice clones virtually indistinguishable from legitimate communications; victims lost $12.5 billion in 2024 alone, up 25% from 2023.
👉 Defense requires AI-powered countermeasures: Organizations must deploy behavioral analytics, multi-factor authentication, metadata verification, and family/corporate authentication protocols to combat AI-generated impersonation attacks before they become undetectable.
👉 The trust crisis is imminent: As video quality improves, experts warn of a "complete lack of trust" in digital interactions becoming society's biggest challenge, forcing fundamental changes in how we verify identity and authenticity across all digital channels.
Fraud just became everyone's problem, something that has long been feared. Deepfake technology has crossed a critical threshold; what once required sophisticated expertise now takes minutes and costs nearly nothing. MIT researchers confirm we've reached "industrial scale" for AI-powered scams, with fraudulent manipulation dominating incident reports for 11 of the past 12 months.
The evidence is staggering: three in ten retail fraud attempts now use AI-generated content, UK consumers lost £9.4 billion in nine months, and deepfake fraud surged 700% in Q1 2025 alone. These aren't isolated attacks - they're systematic campaigns. Scammers impersonate CEOs on video calls to authorize fraudulent wire transfers, clone voices from just three seconds of audio to fool victims into "emergency" payments, and generate fake job candidates who pass technical interviews in real-time.
What makes this shift particularly alarming is accessibility. As Harvard researcher Fred Heiding notes, "It's becoming so cheap, almost anyone can use it now." The technology has shattered traditional barriers to entry, enabling mass-produced personalization that eliminates the spelling errors and awkward phrasing that once flagged scams.
Yet this democratization of deceptive tools also illuminates a path forward. If AI can scale fraud, it can scale defense. Detection firms are already developing behavioral analytics and metadata verification systems. The race isn't between humans and AI scammers, it's between defensive and offensive AI applications.
Why it matters: The industrialization of deepfake fraud fundamentally shifts cybersecurity from an IT concern to an existential business risk, requiring enterprises to rethink identity verification, employee training, and real-time fraud detection systems. As AI tools become universally accessible, the competitive advantage belongs to organizations that deploy AI-powered defenses faster than adversaries can deploy AI-powered attacks.
Sources:
🔗 https://www.theguardian.com/technology/2026/feb/06/deepfake-taking-place-on-an-industrial-scale-study-finds
🔗 https://fortune.com/2026/01/13/ai-fraud-forecast-2026-experian-deepfakes-scams/
Turn AI into Your Income Engine
Ready to transform artificial intelligence from a buzzword into your personal revenue generator?
HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.
Inside you'll discover:
A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential
Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background
Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve
Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.
The $285 Billion Software Shakeup That Changes Everything
Wall Street just got a brutal wake-up call, and AI was holding the alarm clock. Anthropic's launch of Claude Cowork plugins for legal, finance, and marketing workflows triggered a staggering $285 billion selloff across software and data stocks in a single week. Thomson Reuters plunged 18%, LegalZoom dropped nearly 20%, and even giants like Salesforce and Microsoft took heavy hits.
So what happened? Anthropic released open-source plugins that can automate contract review, compliance workflows, and legal research, tasks that thousands of professionals (and expensive software subscriptions) currently handle. Investors panicked, fearing that the entire SaaS business model might be on borrowed time. If AI agents can do the grunt work, why pay per-seat licensing fees?
Here's why this matters to the AI community: We're watching foundation model companies move from building tools for software companies to competing against them. This is the platform layer eating the application layer in real time.
But not everyone is panicking. Nvidia's Jensen Huang called the fears "illogical," and analysts at Wedbush say enterprises won't abandon trillions of dollars in existing software infrastructure overnight. Some beaten-down stocks may actually be bargains now.
This selloff signals that AI is no longer just enhancing software, it's actively threatening the business models that built Silicon Valley. Is this finally the end of SaaS?
Learn how to make AI work for you
AI won’t take your job, but a person using AI might. That’s why 2,000,000+ professionals read The Rundown AI – the free newsletter that keeps you updated on the latest AI news and teaches you how to use it in just 5 minutes a day.