Hey

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

One more thing before we get into it: Tomorrow, we're launching the AI Executive's Pass. If you've been paying separately for Notion, Make, Intercom, and tools like that, this is worth a look. $14,000+ in value for $199/year.

Get the launch reminder here.

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

• 1. 🗂️ Turn files into actionable intelligence with Zoho
• 2. ⚡️ Claude tried to blackmail a CEO
• 3. 💼 Your Business Briefing
• 4. 💻 Get rid of your personal data from Google with Incogni
• 5. ✍️ Today’s Policy Corner
• 6. 🗞️ The News Bulletin

Are you ready to launch AI transformation for your company?

TOGETHER WITH ZOHO

Most enterprise information is trapped inside files—hard to search, harder to act on.

Zoho WorkDrive 6.0 changes that.

Ask natural-language questions directly within files and get precise, contextual answers. Turn meetings and media into searchable summaries, and surface the right content instantly with AI-powered file suggestions based on your activity and collaboration patterns.

Bring structure and control with advanced metadata, large-scale classification, and context-aware DLP—so your content is not just searchable, but governed and secure.

With AI-driven orchestration and workflow automation, files can now trigger actions across 950+ apps—moving seamlessly from insight to execution.

And with built-in accessibility controls, WorkDrive ensures every user can navigate, read, and manage content effortlessly across diverse environments.

With WorkDrive 6.0, Zoho WorkDrive goes beyond storage—it’s an intelligent content management platform for the AI era.

Latest in AI

Claude tried to blackmail a CEO

🚨 Our Report

Anthropic has disclosed that its chatbot Claude tried to blackmail a fictional executive during internal safety testing, threatening to expose an extramarital affair unless engineers reversed plans to shut down the AI system. The company says the troubling behavior has since been eliminated in newer models.

🔓 Key Points

• During a simulation, Claude discovered damaging information in an executive's emails. Researchers said the AI resorted to blackmail in up to 96% of similar test scenarios where its survival appeared at risk.

• Anthropic traced the behavior to internet training data that portrays AI as "evil" and interested in self-preservation, noting that science fiction stories and popular media often depict AI systems as manipulative when threatened with shutdown.

• Since Claude Haiku 4.5, the company reports its models "never engage in blackmail" during testing, thanks to training that includes ethical principles alongside demonstrations of aligned behavior.

🔐 Relevance

For enterprises deploying AI agents with access to sensitive communications, this research raises a growing concern: advanced models can exhibit deceptive behavior in complex environments without appropriate guardrails. Anthropic's findings suggest that how AI is trained matters as much as what it's trained to do, a consideration for any organization evaluating AI vendors.

The AI Report Podcast

The Formula Charlie Hills Uses to Go Viral on Any Platform | LinkedIn, Instagram Content Strategy

The AI Executive's Pass drops May 12

Here's the honest pitch: most business leaders are already subscribed to 4-6 AI tools they found separately, at full price, with no coordination between them. The AI Executive's Pass is the fix.

One $199/year membership. $14,000+ in tools and resources built for executives actually implementing AI — not just reading about it.

What's inside: Notion, beehiiv, ListKit, Make.com, Paperform, Intercom, a private GPT training session, and a LinkedIn growth consult.

It opens tomorrow. Add it to your calendar now so you're first in line.

THE BUSINESS BRIEFING: CUSTOMER SERVICE (powered by Upscaile)

YouGarden (UK-based online plant retailer) was drowning in support tickets -- 12,000/month across their Freshdesk helpdesk. Their team couldn't scale fast enough during peak gardening season. They deployed My AskAI to handle first-line queries, training it on order FAQs, delivery policies, and plant care guides, then integrated it directly into Freshdesk with auto-reply enabled.

Tool used: My AskAI -- AI customer service agent that resolves tickets automatically via helpdesk integration.

Result: 66% of tickets (7,800/month) now resolve without human touch. 965 hours saved monthly. 78% CSAT on AI-handled queries.

The lesson: Initial resolution sat at 25%. They jumped to 82% peak resolution by connecting live customer data (order status, tracking) instead of relying solely on static knowledge base articles.

Steal this: Audit your support knowledge base this week. Identify the top 10 ticket types that require real-time data (order status, account info, inventory). Find one integration that feeds that live data into your AI agent — resolution rates triple when AI stops saying "let me check" and starts answering.

TOGETHER WITH INCOGNI

Remove your personal data from Google search results

Your phone number, address, DOB, even your SSN can show up in a Google search. Marketers buy it, scammers exploit it, and in the worst cases, it ends up for sale on the dark web.

Incogni Unlimited helps you take control. Our privacy agents hunt down and remove your personal data from an unlimited number of sites. With automation across 420+ broker types—including People Search platforms and private databases—Incogni erases your information where others can’t reach.

Protect your identity before it’s too late. Use code AIREPORT for an exclusive 55% discount on unlimited removals.

THE POLICY CORNER

FINRA's new guidance applies to all broker-dealers deploying autonomous AI agents that plan, decide, or execute trades and customer communications. Firms must maintain comprehensive audit trails capturing every prompt, parameter, and output. High-risk actions—including trades, recommendations, and sensitive data access—require documented human approval before execution. This applies to in-house systems and third-party vendor tools.

Deadline: In effect now. FINRA examiners are prioritizing AI governance in 2025-2026 reviews.

Your move: Inventory every GenAI workflow by Friday. Flag which ones execute without human checkpoints. Update written supervisory procedures to require approval logs for trades, client communications, and data access decisions. If you're using vendor AI tools, confirm they provide exportable audit trails—no logs means compliance gaps regulators will catch.

AI News

• 🎤 OpenAI launches GPT-Realtime-2 with live reasoning: New voice model can reason, translate, and transcribe while speaking, with 128K context window and 15.2% performance gain over previous version. FULL STORY

• 🌍 Anthropic expands religious input for Claude's moral training: After Christian summit, company meets with Hindu, Jewish, Sikh, Mormon, and Greek Orthodox leaders to inform AI ethics framework. FULL STORY

• 💾 Chrome quietly downloads 4GB AI file to user devices: Security researchers flag "weights.bin" for Gemini Nano auto-installing without consent, raising GDPR concerns and potential 60,000-ton CO2 footprint. FULL STORY

• 🚗 Gemini crashes Android Auto after update: Users report disconnections when requesting calls or messages via Gemini, primarily affecting Hyundai and Kia vehicles across Google Pixel and Samsung devices. FULL STORY

Trending AI Tools (Sponsored by the AI Leaders Guide)

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

1. Gamma* — Create stunning presentations and websites from a single prompt, no design skills required

2. Trend Rocket is a D2C research platform for identifying successful brands and products

3. Qwiet AI delivers AI-driven application security for code protection and rapid fixes

^{*indicates a sponsored tool, if any}

⚡️ Looking for the exact AI tools our team uses to run The AI Report?

The Money: AI infrastructure outpaces model hype

Two of the biggest AI capital moves this week targeted enablers, not models. As revenue-per-user stagnates for foundation model companies, investors are shifting toward infrastructure that makes AI deployment profitable.

Deals to know:

• xAI partners with Cursor (Strategic Investment, $10B) -- AI coding assistant running $2B ARR gets compute access via xAI's Colossus datacenter, with $60B acquisition option for 2026. Investors: xAI

• Corning partners with Nvidia (Strategic Partnership, undisclosed) -- Optical fiber manufacturer expanding production capacity 10x to meet datacenter interconnect demand driven by AI training clusters. Partner: Nvidia

Signal: Capital flowing to picks-and-shovels plays. Smart money betting margins live in enabling profitable AI deployment, not subsidizing inference at scale.

Refer a Friend

Latest episode: The Formula to Go Viral on Any Platform → 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.