Key points
- Combined AI models and robots can technically handle approximately 81% of US work tasks.
- Robots are currently cheaper than human workers for just 0.3% of all tasks.
- Adding robotics to AI increases task exposure from 50% to 81%, heavily impacting transport roles.
Anthropic’s latest robotics study reveals that artificial intelligence models and physical robots together could technically perform about 81% of the work tasks Americans do today. According to The Next Web, which reported on findings picked up by Business Insider, this marks a significant expansion from earlier research that looked only at AI chatbots. While AI models alone cover roughly half of all work, adding robotics bridges the gap to 81%. This shift primarily affects physical, lower-paid roles that chatbots cannot address, representing a departure from the usual pattern of white-collar AI exposure.
By the numbers
- 81% of US work tasks can technically be handled by AI models combined with robots.
- 0.3% of tasks are currently cheaper to perform with robots than with human labor.
- 90% of transport tasks are exposed to automation when robots are added to AI models.
What the figures mean
The study highlights a stark contrast between technical capability and economic viability. While robots can technically handle about 74% of physical tasks, most of these applications require tightly controlled settings like assembly lines. In the transport sector, exposure jumps from fewer than 15% with AI alone to about 90% with robots included. Nine of the ten most exposed jobs involve vehicle driving, led by taxi drivers. However, cost remains the primary barrier. Currently, only about 300,000 US workers hold jobs where robots are already cheaper, mostly packers and packagers.
What should you do?
Organizations should assess their current operational workflows to identify which physical tasks are technically automatable versus which are economically viable. Teams usually need to review supplier contracts for robotics and AI integration to understand long-term cost structures. It is critical to determine whether internal processes rely on tightly controlled environments that favor automation or dynamic settings that do not. Stakeholders should also evaluate the speed and efficiency of existing automated solutions compared to human performance, as cleaning and maintenance robots are often slower and more expensive.
Operational impact
When evaluating automation reports, leadership must distinguish between technical feasibility and financial reality. Operations teams should audit current physical workflows to identify bottlenecks where robotics might offer speed advantages in controlled environments. Procurement departments need to review vendor claims regarding cost savings, ensuring that initial hardware investments align with long-term operational budgets. Human resources should prepare for potential role shifts in logistics and packaging, focusing on upskilling rather than immediate displacement. Communication with suppliers is essential to understand the maturity of robotic solutions for specific industry tasks.
What to do and how to stay safe: Anthropic
- Review current operational workflows to identify physical tasks that occur in tightly controlled environments suitable for robotics.
- Audit existing automation vendors to compare their cost structures against current labor expenses for specific roles like packing.
- Monitor industry reports for changes in robot pricing, as a 20% drop could significantly increase the number of economically viable automated tasks.
- Evaluate the speed and efficiency of robotic solutions in your sector, noting that some cleaning and maintenance robots are slower than humans.
Step-by-step guide: Shadow AI Explained: 8 Questions Answered for Secure Operations
General security guidance from the BitShield News newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
What percentage of US work tasks can AI and robots technically perform?
According to Anthropic’s study, AI models and robots together can technically handle about 81% of US work tasks.
Are robots currently cheaper than human workers for most tasks?
No, robots are currently cheaper than human labor for only 0.3% of tasks, affecting approximately 300,000 workers.
Which sector sees the biggest jump in automation exposure when robots are added?
The transport sector sees exposure jump from fewer than 15% with AI alone to about 90% when robots are included.
