
Artificial intelligence is supposed to help people finish work faster. Instead of spending hours writing code, checking documents or preparing reports, workers can now ask AI tools to handle much of the work. But what happens when the AI finishes tasks faster than people can review the results?
That is the problem emerging at Anthropic, the company behind Claude, one of the world’s leading AI assistants. The company has reported that its employees are spending more time in meetings, hiring fewer junior technical workers and changing how they handle their daily responsibilities as AI becomes part of almost every stage of their work.
In a report published on October 16, 2026, Anthropic explained how AI is changing the way its employees work and how the company hires. The findings are surprising because they show that AI can help a business produce more while creating new demands on the people responsible for checking, managing and making decisions about that work. Read Anthropic’s report.
Why Are Anthropic Employees Spending More Time in Meetings?
Anthropic found that the average employee now spends about 20% more time in meetings than a year ago. Even technical employees who spend much of their time building software have seen an increase, with meeting time among technical individual contributors rising by 16%.
At first, this may sound strange. If AI can write code, find errors, prepare documents and complete other tasks, employees should have more free time, right? The problem is that completing work is only one part of running a business. People still need to decide what should be done, check whether the results are correct, agree on priorities and take responsibility for important decisions.
AI can generate a large amount of work in a short time. However, employees must still read the results, check for mistakes and decide what to do next. When several AI agents are working on different tasks at once, people can quickly become overwhelmed by the amount of information they need to review.
Meetings help employees discuss these results, settle disagreements and decide which tasks deserve attention. Anthropic says face-to-face coordination remains important because people still need trust, clear responsibility and shared decisions, even when AI handles more of the practical work.
The increase in meetings also reflects the company’s growth. Anthropic is hiring across business functions, and some of those roles naturally require more meetings than technical jobs. Therefore, AI is not the only reason employees are spending more time together, but the company says the change is happening even among technical workers.
Anthropic Is Hiring Fewer Junior Technical Workers
One of the most important findings concerns entry-level jobs. Anthropic has reduced the share of junior employees in its technical departments from 21% in early 2025 to 14% in 2026. The share of new technical hires who are junior employees fell even more sharply, from 27% to just 9%.
These figures matter because junior jobs are often where people begin their careers in technology. A new software developer might start by fixing small errors, writing simple code, testing applications and helping experienced engineers. These tasks allow beginners to gain experience before moving on to more complex responsibilities.
AI coding tools can now handle many of these basic tasks. A more experienced developer can ask an AI assistant to write a piece of code, find a bug or suggest a solution instead of assigning every small task to a junior employee. This can help a team work faster, but it may also reduce the number of entry-level workers a company believes it needs.
Anthropic has not stopped hiring junior workers entirely. The findings show that junior employees make up a smaller share of its technical workforce and new technical hires. The company also continues to hire technical staff, but its hiring in areas such as sales, finance and recruitment has grown faster.
The wider concern is what happens to young people who want to enter the technology industry. If companies use AI to handle the tasks that once gave beginners their first work experience, graduates may find it harder to get their first job. Businesses will need to think carefully about how new workers can gain the skills and experience needed to progress.
AI Agents Are Doing More Work Inside Anthropic
Anthropic’s report also shows how quickly AI agents have become part of its daily operations. An AI agent is a system that can work through a task, use tools and take several steps towards completing a goal, rather than simply answering a question.
According to the report, the number of hours worked by AI agents grew rapidly. By March 2026, agents were working more hours than employees, and by August, the company estimated that agents were working about six hours for every hour worked by a human employee.
This does not mean AI agents are replacing every employee or working in exactly the same way as a person. An hour of AI activity is not directly equal to an hour of human work, because the tasks, quality and results can be very different. However, the figures show how much work Anthropic is now assigning to AI systems.
The company says this change is helping employees produce more code per person, close more sales per sales representative and process more contracts. In simple terms, employees can use AI to complete parts of their work faster and handle more tasks than they could before.
However, Anthropic also recognises that producing more work does not automatically mean producing better work. More lines of code do not guarantee better software, just as processing more contracts does not necessarily mean every contract has been checked perfectly. The quality and business value of the results still matter.
Employees Are Becoming Managers of AI
AI is also changing what employees do during the working day. Instead of completing every task themselves, some workers now spend more time giving instructions to AI systems, checking their work and deciding what should happen next.
Anthropic’s April 2026 employee survey found that 27% of respondents said Claude had made them more productive. At the same time, 20% said their job had shifted from doing the work themselves to managing Claude.
This is a major change in how a job can be organised. An engineer may spend less time writing every line of code and more time deciding what software should do, assigning tasks to AI agents and checking the code they produce. A sales employee may use AI to build a dashboard, organise customer information or prepare reports instead of waiting for a specialist to do all the technical work.
The change creates new responsibilities. Employees need to understand the task, explain it clearly to the AI, review the result and correct mistakes. They also need to know when an AI system should not be trusted to make a decision on its own.
In its report, Anthropic describes human attention and judgement as a growing limit on productivity. AI can generate answers and complete tasks quickly, but people still have a limited amount of time to read, understand and approve the results.
Why More AI Does Not Always Mean Less Work
The experience at Anthropic shows why using AI does not automatically make every working day shorter. AI can reduce the time required to complete individual tasks, but it can also make it possible to start more tasks and produce more work than before.
Imagine a worker who previously spent a full day preparing one report. With AI, the same worker might prepare several reports in that time. This sounds like a clear improvement, but each report still needs to be checked, corrected and shared with the right people. The worker may end up spending the saved time reviewing more reports rather than finishing work earlier.
The same problem can occur when a company uses several AI agents at once. One agent may write code, another may analyse data and a third may prepare a summary. The employee managing them must check whether the results are accurate and whether the separate tasks fit together properly.
This can create a new type of workload. Instead of being limited mainly by how quickly people can perform tasks, the business becomes limited by how quickly people can review results and make good decisions.
That does not mean AI is useless. It means companies need to think beyond speed. They must decide whether the extra output improves customer service, increases revenue, reduces costs or produces better results for the business.
What Does This Mean for Workers and Job Seekers?
Anthropic’s findings offer an important warning for people building careers in technology and other office-based jobs. AI is changing not only how work is done, but also which skills employers may value most.
For junior workers, the challenge is that some basic tasks may become less common as companies rely more on AI. Learning to write simple code or prepare basic reports remains useful, but workers may also need to understand how to use AI tools, check their output and solve problems that require human judgement.
For experienced workers, the ability to manage AI systems could become increasingly important. Knowing how to give clear instructions, spot mistakes, protect sensitive information and connect AI output to business goals can make a worker more effective.
Communication is also likely to remain valuable. When AI handles more of the practical work, employees may spend more time explaining decisions, coordinating with colleagues and agreeing on what the business should do next. These tasks require people to understand each other and take responsibility for the outcome.
However, Anthropic’s experience should not be treated as proof that every company will follow the same path. The company builds AI products and uses its own technology heavily, so its working environment is different from that of a hospital, school, factory or small retail business. Other employers may adopt AI more slowly or use it for different tasks.
What Should Businesses Learn From Anthropic?
Businesses should not measure the success of AI only by how much work it produces. They should also measure whether the work is accurate, whether customers benefit and whether employees can manage the new workload without creating more problems.
First, companies need to identify which tasks AI can handle safely and which require human approval. Routine work such as preparing a first draft or sorting information may be suitable for automation, while sensitive decisions may need careful human review.
Second, managers should make sure employees have time to check AI-generated work. If workers are expected to manage several agents without enough time or training, mistakes may increase and productivity gains may disappear.
Third, businesses should consider how junior employees will learn. If AI takes over basic tasks, companies may need new training programmes that allow beginners to develop skills through supervised projects, practical exercises and more complex assignments.
Finally, companies should review their meeting culture. Some meetings are necessary for decisions and teamwork, but others may simply be the result of unclear instructions or poor communication. Businesses should use AI to remove unnecessary work, not create more discussions about work that could have been handled more simply.
The Bottom Line
Anthropic’s experience shows that AI can make employees more productive while changing the way they spend their time. The company reports that it is producing more code, closing more sales per representative and processing more contracts, but its employees are also spending more time in meetings and managing AI systems.
The decline in junior technical hiring is another important finding. If AI takes over many tasks that beginners once performed, companies will need to find new ways to train the next generation of workers.
The biggest lesson is that AI does not remove the need for people. It changes where people spend their time and what they are expected to do. The companies that benefit most may be those that combine AI’s speed with human judgement, clear communication and good management.
AI was supposed to save time, and in many cases it can. But as Anthropic’s experience suggests, the real challenge is making sure the time saved leads to better work, rather than simply creating more work for people to manage.