Your employees are absolutely using AI and 25% or more are using ChatGPT

It is time to make AI standard issue for anyone who works in front of a computer. It is common to hear that 95 percent of large-scale AI projects fail according to MIT’s 2025 State of AI in Business report. I say SO WHAT. We are focused on the wrong thing and we are not talking about what moves the needle. 

While enterprise AI pilots may struggle, there is a wave of progress happening at the individual level that we aren't really talking about. According to recent OpenAI research (available below), 28 percent of employees are now using ChatGPT for work even if their company does not officially approve or pay for it. Employees are delivering results with or without company support. Is there a security risk? Absolutely. What is your company doing about it? Simply putting a note against AI usage in your policy handbook is not enough! 

When you look across all AI tools, about half of all employees are leveraging AI in some way. That means in any typical workplace two out of every four people are using AI regularly, even if the company does not officially sanction any tools. 

The value is clear. A Federal Reserve Bank of St. Louis study found that over half of frequent AI users save at least three to four hours of work each week. Harvard and MIT found that knowledge workers using generative AI produce work that is up to forty percent higher in quality compared to those who do not use AI. What is the real difference? The 95 percent failure rate applies to bigger enterprise projects using AI. Meanwhile, the real opportunity is to start small. Everyday tasks can lead the way to bigger and much more valuable AI automation. Hit me up! What do you see happening at your company? 

See more details of the study here: OpenAI Study 

#ArtificialIntelligence, #AI, #GenerativeAI, #AIForBusiness, #FutureOfWork, #Productivity, #WorkplaceInnovation, #Innovation, #Technology, #Leadership, #ChangeManagement, #DigitalTransformation, #AIAutomation, #ResponsibleAI, #AIAdoption,#ChatGPT

AI’s 95% Failure Rate Looks Just Like Cloud’s Early Struggles

**AI’s 95% Failure Rate Looks Just Like Cloud’s Early Struggles** MIT says 95% of AI pilots flop. Sounds scary, right? But this is not the end of the world. In fact, it looks a lot like the early days of cloud computing. Fifteen years ago, most cloud projects went nowhere. Security fears, messy migrations, and grumpy IT teams kept things stuck in pilot mode. Today, AI pilots are tripping on the same issues. Poor workflow fit, bad data, and unrealistic hopes keep projects from scaling. Here is the good news. Cloud eventually became the backbone of business. AI will get there too and probably faster since it is standing on cloud’s shoulders. So yes, most pilots fail. Think of them as practice rounds. The companies that stick with it and learn from the flops will be the ones cashing in when AI finally clicks.

Microsoft Pushes Copilot Deeper Into Everyday Work

# Microsoft Pushes Copilot Deeper Into Everyday Work Microsoft is bringing Copilot everywhere. Starting in October the Copilot app will automatically appear on Windows devices with Microsoft 365 apps like Word, Excel and PowerPoint. In Power BI it is turned on by default. In Windows 11 it is being built deeper into the system. In retail it is being tested as a shopping assistant. This is what adoption looks like. Not a big launch. Not a dramatic shift. Just a steady blending of AI into the tools people already use. And it leads to a simple question. Will people even notice. Will they know when it is Copilot formatting, writing, or analyzing. Or will it feel no different from normal software doing its job. That may be the real turning point. AI is not something you start using one day. It is something that slowly becomes part of everything you already use. **Labels:** Microsoft, Copilot, AI Adoption, Windows 11, Microsoft 365, AI Integration

The Global AI Divide: Can Cell Phones Help Close the Gap?

The Global AI Divide: Can Cell Phones Help Close the Gap?

Newly released reports today on AI adoption do not surprise me. They show that clearly wealthier countries are moving fast while developing nations are being left behind. It reminds me of the early spread of the internet. At first, only a few places had real access, and it widened the gap before mobile phones helped bring more people online.

This makes me wonder about AI. Could cell phones play the same role again? Most people in the world may never own a powerful computer, but billions already have a phone in their pocket. If AI tools are built to run well on mobile devices, access could spread much faster. The question is whether large language models can become small enough and efficient enough to run locally on phones. With the right advances in chips and model design, this could become reality.

The big question is whether new phones and networks will make this possible. Will the next generation of mobile technology help close the divide, or will AI stay concentrated in the wealthiest nations? The answer will shape who really benefits from this revolution. **Labels:** global AI trends, AI adoption, Cell Phones, Local AI, AI inequality, mobile AI, digital divide