Connect with us

Hi, what are you looking for?

AI Research

Anthropic Study Reveals AI Could Replace 94% of Computer Jobs, Automation Remains Limited

Anthropic’s study reveals AI could automate up to 94% of computer jobs, yet current implementation lags significantly, with only 33% in practice.

Artificial Intelligence (AI) is reshaping the landscape of employment as companies increasingly adopt AI tools to boost efficiency. This shift has triggered concerns among employees about job security. A recent study conducted by **Anthropic** offers insights into which jobs are most susceptible to automation by AI and which roles appear to be more resilient for the time being.

The study, titled “Theoretical Capability and Observed Usage by Occupational Category,” examined nearly 2 million interactions with the AI model **Claude**. It juxtaposed AI’s potential for automation against its current applications in the workforce. The findings reveal a significant discrepancy between the theoretical capabilities of AI and its observable deployment in various occupational settings.

According to the research, occupations in the computer and mathematics fields exhibit the highest theoretical vulnerability to AI, with an exposure rate of approximately **94%**. Close behind are office and administrative roles at **90%**, and legal professions also demonstrate considerable risk. Other sectors, including business, finance, architecture, engineering, and management, all report exposure levels exceeding **60%**.

However, the actual implementation of AI in workplaces remains significantly lower than these theoretical estimates. In fact, the highest observed automation rate is **33%** for computer and mathematics positions, while many sectors register less than **20%** in terms of AI usage. Physical labor roles, such as those in agriculture, construction, and grounds maintenance, show virtually no signs of automation, underscoring a gap between potential and practical application.

Specific job roles are already feeling the impact of AI. For instance, computer programmers exhibit an observed exposure of around **75%**, while data entry operators are at **67%**. The study also highlighted a trend of slower hiring for younger workers in industries that are particularly vulnerable to AI disruption.

Experts caution that AI has not yet reached its full potential and as the technology matures, the degree of automation is likely to escalate in the coming years. This evolving landscape emphasizes the importance of **upskilling** and adapting to AI tools. Workers will need to enhance their skill sets to remain competitive in a job market increasingly influenced by AI advancements.

The findings of this study serve as a crucial reminder of the ongoing transformation within the workforce driven by AI technology. As companies continue to leverage AI for operational efficiency, the dynamic between technological advancement and job security will remain a pivotal issue for both employers and employees alike.

See also
Staff
Written By

The AiPressa Staff team brings you comprehensive coverage of the artificial intelligence industry, including breaking news, research developments, business trends, and policy updates. Our mission is to keep you informed about the rapidly evolving world of AI technology.

You May Also Like

AI Government

US government mandates AI firms like Anthropic grant irrevocable “any lawful use” licenses for federal contracts amid rising scrutiny and procurement standards.

AI Government

Albanese government warns OpenAI and Anthropic to align AI development with Australian values or face a decade of strict regulations and penalties.

Top Stories

Amazon, Google, and Microsoft continue to support Anthropic AI despite Pentagon risk labels, emphasizing their commitment to AI innovation amid regulatory challenges.

AI Generative

All major LLMs, including OpenAI's GPT series, showed significant potential for academic fraud, with Grok-3 facilitating misconduct over 30% of the time.

AI Regulation

A bipartisan coalition has unveiled the Pro-Human Declaration, urging a moratorium on superintelligence development until safety consensus is achieved, reflecting 95% public opposition to...

AI Research

Anthropic and ETH Zurich reveal AI can deanonymize online accounts with a remarkable 68% recall rate, raising privacy concerns across digital platforms.

AI Regulation

Anthropic's AI model Claude faces contract cancellation with the DoD over surveillance restrictions, raising urgent concerns about privacy and governance.

AI Government

Anthropic contests its classification as a supply chain risk by the Pentagon, asserting its AI model Claude won't support mass surveillance or autonomous weapons.

© 2025 AIPressa · Part of Buzzora Media · All rights reserved. This website provides general news and educational content for informational purposes only. While we strive for accuracy, we do not guarantee the completeness or reliability of the information presented. The content should not be considered professional advice of any kind. Readers are encouraged to verify facts and consult appropriate experts when needed. We are not responsible for any loss or inconvenience resulting from the use of information on this site. Some images used on this website are generated with artificial intelligence and are illustrative in nature. They may not accurately represent the products, people, or events described in the articles.