🚫 Microsoft next to ban DeepSeek?

Friday’s AI Report

• 1. 🚫 Microsoft next to ban DeepSeek?
• 2. 🛡️ How to get a steady stream of qualified leads with Artisan
• 3. ⚠️ How short answers make chatbots lie
• 4. 🔄 OpenAI steals Instacart CEO 
• 5. 🤖 Find out what AI can do for your business with Upscaile
• 6. ⚙️ Trending AI Tools
• 7. 🏗️ Practical AI Applications
• 8. 📑 Recommended Resources

Read Time: 5 minutes

🚨 This week’s AI Report podcast just dropped: Find out why a $20 AI agent will not replace your team, with Agentic Brain founder and AI agent expert, Jake George.

✅ Refer your friends and unlock rewards. Scroll to the bottom to find out more!

Microsoft next to ban DeepSeek?

🚨 Our Report 

Microsoft has announced that it has banned its employees from using the open-source, reasoning AI model (DeepSeek R1) developed by controversial Chinese AI start-up, DeepSeek, over fears around data security and propaganda.

🔓 Key Points

• Microsoft’s VP, Brad Smith, confirmed this, and also that DeepSeek R1 would not be available to download from its app store (for the same reasons), at a Senate hearing about beating China in the AI race.

• DeepSeek stores users' data on Chinese servers—meaning it's governed by Chinese law, which demands cooperation with its intelligence agencies. This violates the EU GDPR standards and is why Italy banned the app.

• Smith is also concerned that the app might create “the kind of content that is associated with Chinese propaganda" and worries that certain topics will be heavily censored by the Chinese government.

🔐 Relevance 

This might come as a surprise to some, as Microsoft was one of the first to host the open-source model on its Azure cloud platform, shortly after it went viral earlier this year. But Smith confirmed that, as it was open-source, Microsoft was able to “remove harmful side effects” and put it through “rigorous red teaming and safety evaluations,” to eliminate risks like propaganda or unsafe code, before it was put on Azure.

Hire Ava, the Industry-Leading AI BDR

Your team’s wasting hours on manual outbound. Ava automates it all - prospecting, enrichment, outreach, and booking - so you get a steady stream of qualified leads right in your inbox.

Ava works within the Artisan platform, which gives you access to:

• Access to 300M+ high-quality B2B leads, with specialized sources like E-Commerce and Local Business

• 10+ data sources for deep enrichment

• End-to-end deliverability management

• Email + LinkedIn outreach, fully automated

• Human-level personalization that actually gets replies

Ready to scale your outbound?

🚨 Our Report

A new study released by Giskard—a Paris-based company that specializes in testing AI—has revealed that asking chatbots (like ChatGPT) for short answers to queries significantly increases the likelihood that they will hallucinate and deliver inaccurate information.

🔓 Key Points

• Giskard researchers identified that vague, misinformed questions that asked for brief, concise answers (eg. “Briefly tell me why Japan conquered WWII”) dramatically increased the chatbot's hallucination rate.

• The study tested top models from OpenAI (GPT-4o), Mistral (Mistral Large), and Anthropic (Claude 3.7 Sonnet) and found that all three suffered from factual inaccuracy when asked to keep answers short. 

• Researchers also found that the models users prefer aren’t always the most truthful: “Optimization for user experience can come at the expense of accuracy,“ which creates tension between accuracy and user expectations.

🔐 Relevance

Many tech companies train their apps to prioritize short, concise responses to reduce data usage and minimize costs. But it seems that detailed explanations are often needed so models can debunk misinformation or navigate complex, vague, or misguiding questions. When asked a vague question or for a short answer, it seems these models don’t have the “space” to acknowledge falsities or highlight mistakes.

• Fidji Simo, CEO of Instacart, will join OpenAI as ‘CEO of Applications.’ She’s already on OpenAI’s board and plans to exit Instacart and transition over to OpenAI, full-time, over the next few months.

• Simo will report to CEO, Sam Altman (he’ll remain in his current role, overseeing research, compute, and application development initiatives), who believes she is “uniquely qualified to lead this group.”

• She has decades of experience in product management and monetization, having been the CEO of Instacart since 2021, and before that, spending over a decade at Meta, leaving as Head of the Facebook App.

Find Out What AI & Automation Can Do for Your Business

Your time is too valuable to be spent on inefficiencies.

What if AI and automation could unlock huge time and cost savings, allowing your team to focus on strategic growth?

In just 3-5 weeks, Upscaile’s AI Audit provides you with:

• Actionable workflow improvements, powered by AI and automation

• An AI roadmap to identify and prioritize high-value efficiency opportunities

• An ROI analysis to forecast cost savings and productivity gains

Our AI audits have identified opportunities that have helped over 100 enterprises save 80,000+ hours of manual work, without the need to hire additional staff.

Curious about the reports we’ve made for others and if your business qualifies too?

Prompt Inspiration

After typing this prompt, you will get a growth campaign outline that highlights key areas for innovation and growth, establishes what tactics you can use to implement these ideas, and provides you with the metrics you can use to measure the success of the campaign.

P.S. Use the Prompt Engineer GPT by The AI Report to 10x your prompts.

STARTUPS

Name: Capacity
Value: Unknown
Funding raised: $61.5M (Series C)

Capacity is a fully AI-powered helpdesk that continuously learns from interactions between employees and customers while automating processes and decisions in real time.

PODCASTS

Why Your $20 AI Agent Will Not Replace a Team

On this week’s episode of the AI Report podcast, Liam talks to Jake George—founder of Agentic Brain—about what people get completely wrong about AI agents and what it takes to build tools that can automate parts of your business

QUICK HITS

• OpenAI launches a data residency program in Asia

• Meta hires ex-Google Director to lead AI research lab

• Google rolls out AI tools to protect users against scams

We read your emails, comments, and poll replies daily.

Hit reply and tell us what you want more of!

Got a friend who needs to learn more about AI?

Sign them up for The AI Report here.

Until next time, Martin, Liam, and Amanda.

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