The right way to use AI at work

illuminem summarises for you the essential news of the day. Read the full piece on Fast Company or enjoy below:

🗞️ Driving the news: A new Stanford University study reveals that most companies are using artificial intelligence in the workplace incorrectly — focusing on the wrong tasks and failing to align technology with human preferences
• The research, conducted by Stanford’s Institute for Human-Centered AI and Digital Economy Lab, analyzed the daily work patterns of 1,500 American employees across 104 professions and found that AI tools often target areas where people don’t want automation or where AI still performs poorly

🔭 The context: Despite over $100 billion in AI spending by major U.S. tech firms this year, only 14% of large non-tech companies have meaningfully deployed AI, according to Goldman Sachs
• Stanford researchers identified four categories of work tasks — “Green Light,” “Yellow Light,” “Red Light,” and “No Light”—to determine where AI can be most effective
• The study found that 41% of current AI tools focus on tasks workers want to keep human-led, while only a small share target the “Green Light Zone,” where AI can perform repetitive, low-value tasks efficiently and with user support

🌍 Why it matters for the planet: Although not directly environmental, the findings have sustainability relevance
• Misapplied AI wastes vast computational and energy resources, while well-implemented AI can streamline workflows, reduce inefficiencies, and lower digital carbon footprints
• Aligning AI development with human needs and real-world tasks could foster a more balanced, human-centric, and resource-efficient digital economy

⏭️ What's next: Stanford’s researchers recommend that businesses focus on collaboration over replacement, using AI to automate “the boring bits” — scheduling, documentation, data entry — while maintaining human oversight for judgment-based or creative work
• The study predicts a wave of next-generation AI tools designed for human-AI partnership rather than full automation
• For workers, reframing AI as a collaborator rather than a competitor will be key to long-term productivity gains

💬 One quote: "If you try to use AI to replace your entire job, you’ll probably fail. But if you focus on automating the parts you hate, you’ll thrive." — Stanford AI Research Team

📈 One stat: 69.4% of workers said they want AI to “free up time for higher-value work”, while only 35.6% support fully autonomous AI systems

See on illuminem's Data Hub™ the sustainability performance of AI companies like Anthropic, OpenAI, and Google

Click for more news covering the latest on corporate governance