# **SEO-Optimized Article: Anthropic CEO Claims AI Hallucinates Less Than Humans – Here’s Why**
## **Meta Description:**
Anthropic CEO Dario Amodei argues that AI models hallucinate less than humans but in more surprising ways. Discover the latest Claude 4 updates and the future of AGI.
## **H1: Anthropic CEO Claims AI Hallucinates Less Than Humans – Here’s Why**
### **H2: AI Hallucinations vs. Human Errors**
During Anthropic’s *Code With Claude* event, CEO Dario Amodei made a bold statement: **AI models hallucinate less than humans**, though their mistakes are often more unexpected.
Key takeaways from his discussion:
– **AI hallucinations** occur when models generate false but confident responses.
– **Humans make mistakes regularly**—politicians, journalists, and professionals frequently misstate facts.
– **AI errors stand out more** because they deviate from expected behavior.
### **H2: Anthropic’s Latest AI Advancements**
At the event, Anthropic unveiled **two new Claude 4 models** with enhanced capabilities:
– **Improved memory & reasoning**
– **Better coding & writing performance**
– **State-of-the-art (SOTA) scores** – Claude Sonnet 4 achieved **72.7% on the SWE-Bench coding benchmark**
### **H2: The Road to Artificial General Intelligence (AGI)**
Amodei believes hallucinations **won’t stop AI from reaching AGI**—a system that can learn and perform tasks without human intervention. He even suggested **Anthropic could achieve AGI by next year** (as per a 2024 paper).
### **H2: AI’s Legal Blunders – A Recent Example**
Despite advancements, AI still makes mistakes:
– Anthropic’s **lawyer had to apologize** in court after Claude cited a non-existent legal case.
– The error occurred in an ongoing **copyright lawsuit** involving song lyrics.
### **H2: The Future of AI – Challenges & Potential**
While AI hallucination remains a concern, Amodei argues that **human errors are far more frequent**. The key difference? AI’s mistakes are **more noticeable and sometimes bizarre**.
### **Final Thoughts**
As AI evolves, so does its ability to reason accurately. However, **balancing intelligence with reliability** remains a challenge.
—
**Follow for more tech updates:**
– **X (Twitter), Facebook, WhatsApp, Threads, Google News** – @Gadgets360
– **YouTube** – Subscribe for the latest tech videos
– **Who’sThat360** – Follow for influencer insights
Would you trust AI over human judgment? Let us know in the comments! 🚀
Related: Overwatch 3 rumors
Related: Best gaming backpack 2025
Also read: SEMRush
Also read: DeepSeek