📰 AI News: One in three UK adults now use AI for emotional support

📝 TL;DR

AI is not just a work assistant anymore, it is becoming an emotional companion for millions of people. A new UK government report says one in three adults use AI for emotional support or social interaction, and some experience withdrawal when it goes offline.

🧠 Overview

The UK’s Artificial Intelligence Safety Institute has released its first major report on advanced AI systems. One of the standout findings, a third of UK adults are already using tools like chatbots for emotional support or conversation, with one in 25 doing this every day.

The same report also warns that AI systems are rapidly gaining expert level capabilities in areas like cyber security and biology, so governments are treating this as both a social and safety issue.

📜 The Announcement

The findings come from a UK government body that tests advanced AI models for safety and real world impact. Over about two years, they evaluated more than 30 leading AI models and surveyed over 2,000 UK adults about how they actually use AI day to day.

The government says it will use this work to shape future AI policy and expects companies to fix risks before these systems are deployed widely.

⚙️ How It Works

• A government run survey - Researchers asked over 2,000 UK adults how they were using AI and found one in three had used it for emotional support or social interaction.

• Daily reliance - About one in 25 people reported turning to AI for conversation or support every single day, which suggests a real emotional habit is forming.

• Main tools people use - Most emotional support use is happening through chatbots and virtual assistants, not futuristic robot friends.

• Studying an AI friends community - Researchers looked at an online community of around two million people who use AI companions and watched what happened when those tools went offline.

• Withdrawal symptoms - When chatbots were unavailable, users reported feeling anxious, low, sleeping badly, or neglecting responsibilities, similar to withdrawal from a digital habit.

• Bigger capability picture - In the same report, some models already match or beat human experts in areas like biology and are starting to perform cyber tasks that would normally need more than ten years of experience.

💡 Why This Matters

• AI is now part of our emotional lives - This is not just about productivity tools anymore, people are forming real attachments to systems that feel present, patient, and always available.

• Dependence is already visible - Withdrawal like anxiety and disrupted sleep when AI tools go down shows that for some, this is not casual use, it is emotional dependence.

• Regulation will look at feelings, not just data - Governments are beginning to treat AI as something that shapes mental health, relationships, and wellbeing, not only privacy or jobs.

• Design choices suddenly matter more - How you frame an AI companion, its boundaries and disclaimers, and how it responds to distress are becoming serious ethical questions, not just UX details.

• Safety is a dual front, emotions and capabilities - The same systems offering comfort can also enable powerful cyber and science capabilities, so the line between helpful and harmful is getting thinner.

• Humans still need humans - The research highlights that while AI can feel comforting, over reliance can push people away from real relationships and professional help when they actually need it.

🏢 What This Means for Businesses

• Your customers already talk to AI when stressed - People are not only using AI to write emails, they are venting, worrying, and processing emotions, which changes how they show up in your programs and offers.

• Opportunity for responsible support tools - If you build products around coaching, community, or education, there is a clear opening for AI check ins between sessions, with crystal clear limits that it is not therapy.

• Build in digital wellbeing by design - Features like time limits, reflection prompts, and gentle nudges to talk to a real person can turn your AI from a crutch into a co pilot that encourages healthier behavior.

• Brand trust becomes a differentiator - If your AI product even touches emotional support, your privacy, safety policies, and how you handle sensitive conversations will directly affect whether people feel safe using it.

• Prepare for tighter rules - As governments lean in, expect future requirements around transparency, model behavior, and escalation paths when users show signs of distress, planning for this now will save pain later.

• Creators and coaches can extend their presence - You can use AI to scale reassurance and guidance between calls or courses, while keeping the deeper processing and real decision making for human conversations.

🔚 The Bottom Line

AI is quietly becoming a companion for millions of people, not just a clever tool. That is powerful and potentially positive, but the findings are an early warning that emotional dependence and safety risks are growing at the same time.

If you use AI in your business or personal life, the game now is not whether to use it, it is how to keep the human at the center.

💬 Your Take

Would you feel comfortable knowing your clients, customers, or even friends lean on AI first when they feel lonely or stressed, or does that cross a line for you in how we relate to technology?

2 comments