{"id":"f-73cd9b4b27-ec485726-a1ae-5950-bbf4-6e43d1870a9e","title":"New Anthropic study finds AI could automate many tasks but real workplace use remains limited","description":"A new study by Anthropic finds artificial intelligence is far from reaching its full job automation potential. While roles like programmers and customer service workers show high exposure to AI tools, researchers say there is still little evidence of large scale job losses so far.","timestamp":1772786658,"category":"technology","headline":"New Anthropic study finds AI could automate many tasks but real workplace use remains limited","modifiedHeadline":"AI Automation Potential Exceeds Real Use","subCategories":["scid_b7e786a9"],"subcategoriesName":["Artificial Intelligence"],"region":"IN","bundleMetaData":{"bundleId":"c35cfc57-1bba-4539-a24c-695ea99ac023","bundleTitle":"Mastering AI","cardData":null},"iabCategoryId":"597","iabCategoryName":"Artificial Intelligence","fcSummary":"a new study by anthropic reveals a gap between ai's theoretical capabilities and its current workplace application. while ai could automate many tasks, actual usage remains limited, particularly in computer and math-related fields. occupations like programming and customer service show high exposure, potentially slowing job growth slightly, but widespread job losses aren't yet evident. hiring for ai-exposed roles has dipped among younger workers.","summary":"AI's job automation potential is vast, but real-world use lags. Study finds limited impact on jobs so far, despite high exposure in some roles.","ctaUrl":"https://trends.glance.com/newz/astrology/IN/en/news9/f-73cd9b4b27-ec485726-a1ae-5950-bbf4-6e43d1870a9e","sourceUrl":"https://www.news9live.com/technology/artificial-intelligence/anthropic-ai-labor-market-impact-study-jobs-2939847","bannerImage":"https://g-mob.glance-cdn.com/public/fc/image/pvLB76e7Lm40.webp","content":"

New Delhi: Artificial intelligence is changing how people work. Yet the impact on jobs may be slower and more uneven than many early predictions suggested.

\n\n

A new study by AI company Anthropic attempts to measure how large language models are actually affecting the labour market.

The research introduces a new metric called “observed exposure”. It combines two things. First, the theoretical ability of AI tools to perform certain tasks. Second, how people are actually using AI systems like Claude in real work settings. The results show a large gap between what AI could do in theory and what it is currently doing in workplaces.

\n\n

AI still far from its full job automation potential

Anthropic researchers found that artificial intelligence has not yet reached its full potential in automating jobs. Many tasks that AI could technically perform are still handled by humans.

\n\n

The report explains that “AI is far from reaching its theoretical capability: actual coverage remains a fraction of what’s feasible.”

In simple words, the technology could do much more. It just has not spread everywhere yet.

The study used data from several sources, including occupational task databases and real usage of AI systems. Researchers looked at nearly 800 occupations and the tasks workers perform daily.

A comparison between theoretical AI capability and real world usage shows a clear pattern. Computer and math-related jobs have the highest theoretical exposure. Yet actual AI use in those jobs remains much lower.

\n\n

Jobs most exposed to AI automation

The study also identified occupations where AI tools are already automating parts of the work.

According to the data, the most exposed roles include

• Computer programmers
\n • Customer service representatives
\n • Data entry workers
\n • Medical record specialists
\n • Market research analysts
\n • Sales representatives
\n • Financial analysts
\n • Software quality testers
\n • Information security analysts
\n • Computer support specialists

Computer programmers ranked highest. About 74.5 per cent of their work tasks show exposure to automation in the study’s model. Customer service representatives follow with around 70.1 per cent exposure. Data entry roles show 67.1 per cent.

\n\n

Many of these jobs involve digital information processing. That makes them easier for AI systems to assist or automate.

AI exposure linked to slower job growth projections

Researchers also compared AI exposure with employment forecasts from the US Bureau of Labour Statistics.

The findings suggest a mild connection. Jobs with higher AI exposure may grow slightly slower over the next decade.

The report notes that for every 10 percentage point increase in AI task coverage, projected employment growth falls by about 0.6 percentage points.

Still, there is no clear sign of widespread job losses.

\n\n

Researchers wrote that they found “limited evidence that AI has affected employment to date.”

Early signals appear in hiring patterns

One interesting signal appeared among younger workers.

The research suggests hiring into AI exposed jobs may have slowed slightly for people aged 22 to 25. Data shows a drop of about 14 percent in job finding rates for younger workers entering highly exposed occupations after the launch of ChatGPT.

At the same time, unemployment rates have not increased for these groups.

For now, the message from the study is cautious. AI is spreading quickly. Its real impact on employment may take years to become visible.

\n\n

As the report notes, the goal is to track changes over time. The authors say future updates may reveal clearer economic effects as AI adoption grows across industries.

","readDurationInMinutes":"6","author":{"name":"News9","logo":"https://glance-mob.glance-cdn.com/public/bubble/png/0e26af53-6a49-4ee1-a63d-25161b714ad5.jpeg"},"ism":false,"analyticsResponses":[{"measurementId":"G-EQBYL0ZYL8","type":"GA"},{"measurementId":"33425927","type":"COMSCORE","partnerName":"glance"}],"pointerSummary":["Anthropic’s study finds AI’s job automation potential is unrealised, with a gap between theory & practice.","Computer programmers top the list (74.5% exposure), followed by customer service & data entry roles.","AI exposure may slightly slow job growth, but no widespread job losses are seen; future updates are planned."],"tags":["AnthropicStudy","AIAutomation","JobMarket","LaborMarket","ArtificialIntelligence","LLMs","LargeLanguageModels","WorkplaceImpact","JobLosses","JobExposure","ComputerProgrammers","CustomerService","DataEntry","ChatGPT","EmploymentTrends","USBureauOfLaborStatistics","EconomicImpact"],"redirectionEnabled":false,"consent":true}

New Anthropic study finds AI could automate many tasks but real workplace use remains limited

SUMMARY