Hey

Welcome to The AI Report — the #1 AI newsletter for business leaders.

You don’t need to be technical. Just informed.

• 1. 💪 Create the smoothest presentations with Templafy
• 2. ⚡️ Trump delays OpenAI's GPT-5.6
• 3. 💼 Your Business Briefing
• 4. 👨‍🏫 Teach your whole team AI with Upscaile
• 5. ✍️ Today’s Policy Corner
• 6. 🗞️ The News Bulletin

Are you ready to launch AI transformation for your company?

TOGETHER WITH TEMPLAFY

The enterprise-grade AI behind 4M+ professionals' presentations. Now free for everyone.

Templafy powers document creation for over 4 million professionals at the world's biggest companies. Now, the same enterprise-grade AI is available as a free prompt-to-PowerPoint agent.

Describe what you need. Pick a theme (light, dark, or neutral). Get a polished, structured deck in minutes. Download the .pptx. Open it in PowerPoint. Edit it however you want. No credit card. No trial. No app to install.

This is the presentation quality that Fortune 500 teams rely on, and it costs you nothing.

Latest in AI

Trump delays OpenAI's GPT-5.6

🚨 Our Report

OpenAI will not release its newest model, GPT 5.6, to the public as planned. Instead, the Trump administration has directed the company to distribute it only to a select group of approved partners, with government agencies reviewing access "customer by customer" before any broader release occurs, according to The Information.

🔓 Key Points

• CEO Sam Altman told staff this week that the Office of the National Cyber Director and the Office of Science and Technology Policy "worked closely" with OpenAI on the release plan, which includes a limited preview period followed by a potential general release "a couple of weeks later."

• The move follows an executive order signed earlier this month directing AI companies to voluntarily submit new models to the government for testing and evaluation before public deployment, marking a shift from the administration's earlier "hands off" AI stance.

• The restriction mirrors Anthropic's Project Glasswing approach, where it limited access to its Claude Mythos cyber model to select partners, citing concerns that frontier AI models capable of identifying and exploiting software vulnerabilities at scale could enable unprecedented cyberattacks.

🔐 Relevance

For enterprise buyers, this signals a new reality: advanced AI capabilities will increasingly face government gatekeeping before reaching commercial markets. Organizations planning deployments around frontier models should anticipate approval delays and restricted access periods. The shift also suggests regulators view cyber-capable AI as a national security concern, which could shape future procurement, compliance, and vendor selection requirements.

The AI Report Podcast

The Internet Is Becoming More Centralized. Here's Why It Matters | Ajit Varma, Firefox

THE BUSINESS BRIEFING: HR & TALENT (powered by Upscaile)

Ambea (Nordic care provider, 41,000 employees) processed 100,000+ job applications annually for 3,000 hires across Sweden, Norway, Denmark and Finland—supported by a single central recruiter. Hiring managers ran entire care facilities and couldn't spend days sifting CVs. Manual screening meant hundreds of applications per role, inconsistent evaluation, and strong candidates slipping through.

Tool used: Hubert — AI-powered candidate screening integrated directly into ATS (Teamtailor), delivers structured competency-based interviews and ranked shortlists.

Result: 74% reduction in screening activity. ~3 hours saved per hire. 10/10 candidate experience ratings from majority of applicants. Faster time-to-hire. One recruiter now supports 3,000 annual hires handling 100,000+ applications.

The lesson: AI screening only scales when it sits inside your existing workflow. Ambea integrated Hubert directly into their ATS—no separate platform, no manual exports. Hiring managers see ranked shortlists in the same system they already use. Friction kills adoption.

Steal this: Audit your ATS this week for native AI screening integrations. If your current system doesn't support them, flag it to your tech team. Tools that live outside your hiring workflow create bottlenecks, not efficiency.

TOGETHER WITH UPSCAILE

Teach your whole team AI — the Stanford way

Right now a few people are experimenting with AI, the rest are hesitant, and there's no shared standard.

Upscaile fixes that in 90 minutes.

It's the same curriculum Arturo Ferreira delivered to 50+ Stanford students, refined across 12+ enterprise teams.

One live virtual session, unlimited attendees, $5,000 flat — while typical workshops charge $500–$1,500 a seat. Your team leaves with prompting frameworks, safe repeatable workflows, and AI applied to your real use cases, plus all the materials to keep.

THE POLICY CORNER

Seven major tech firms fined over $3.5 billion for AI training violations since 2022 — Anthropic leads at $1.5B

Regulators across the US and EU have levied $3.5 billion in penalties against Anthropic, Meta, Google, OpenAI, Amazon, Apple, and Clearview AI for training AI on copyrighted works, biometric data, and personal information without consent. Anthropic's $1.5 billion settlement for using pirated books and Meta's $1.4 billion penalty for facial recognition data training represent 81% of total enforcement. Nine of ten cases centered on unauthorized data use — only Apple's $250 million settlement addressed overpromising capabilities. Clearview AI has ignored $105 million in EU fines, claiming no European jurisdiction.

Deadline: Enforcement active now across multiple jurisdictions.

Your move: Audit your AI vendor contracts this month — verify legal basis for training data, confirm GDPR compliance if operating in EU, and document consent chains for any proprietary or user-generated content feeding models.

AI News

• 🏆 Claude gains ground on ChatGPT among paid consumers: Credit card data shows Anthropic's paying consumer base up 75% since January 2026, with interest in Claude courses surging 18x in 30 days. FULL STORY

• 📊 California launches first AI job-loss tracker: New dashboard monitors unemployment claims in AI-exposed occupations statewide, showing no evidence of mass displacement yet but sustained increases among Bay Area tech workers. FULL STORY

• 👥 DeepSeek plans to double headcount across all departments: Chinese AI startup announces aggressive hiring push for core R&D roles following $7.36B funding round at $50B+ valuation. FULL STORY

• 💼 77% of US small businesses now use AI regularly: Intuit data shows adoption jumped from 48% in mid-2024, with marketing (45%), customer service (37%), and bookkeeping (35%) as top use cases. FULL STORY

Trending AI Tools (Sponsored by the AI Executive’s Pass)

A curated look at the AI tools quietly transforming how teams work.

1. Kit:* 600K+ creators. 2.5B emails monthly. 99.8% deliverability. The newsletter OS for creators.

2. NoowAI is an AI virtual assistant for customer support through natural language processing

3. 0PTIKUBE is a visualization and optimization platform for Kubernetes clusters

^{*indicates a sponsored tool, if any}

⚡️ Why pay for 4–6 separate AI tools at full price when the AI Executive’s Pass fixes that?

The Money: Agent evaluation infrastructure

Two of the largest AI funding rounds this week went to companies building the testing layer for autonomous agents—not the agents themselves. As labs race to ship agentic systems, investors are betting the bottleneck is verification, not capability.

Deals to know:

• Patronus AI (Series B, $50M) — Simulated digital environments that stress-test AI agents across multi-day workflows. Revenue up 15x YoY. Investors: Greenfield Partners, Notable Capital, Lightspeed, Datadog, Samsung

• Trase (Seed, $107M) — AI operating system for regulated industries, connecting data sources and foundation models for healthcare and defense agents. Investors: Arch Venture Partners, Red Cell Partners

Signal: Before agents book your flights or run financial analysis, someone has to prove they won't fail. Testing infrastructure now commands premium seed valuations.

Refer a Friend

Latest episode: The Internet Is Becoming More Centralized. → Listen here
Want to reach 400k+ decision-makers? → Sponsor us

Until next time, Arturo and Liam.

P.S. Unsubscribe if you don’t want us in your inbox anymore.