Anthropic Cowork AI Assistant sent a shockwave through Wall Street this week. Today, Anthropic is taking another step forward by improving its model.
Anthropic’s new Claude Opus 4.6 model, announced Thursday, is designed to improve Cowork AI for office and coding work, which could raise even more concerns that the AI tool could replace the specialized software packages that companies use for these tasks.
Legal and financial analysis software stocks have plunged in recent days, dragging the broader stock market with them. The Nasdaq just experienced its worst two-day fall since April, and it was down another 0.7% on Thursday.
Many experts wonder whether AI will ultimately cost some workers their jobs. Tech giants like Anthropic, OpenAI and Google are in a race to create AI models that they hope will support the future workplace.
The anthropogenic rival OpenAI has just been presented a new platform to create AI agents intended to function as colleagues on Thursday morning. And Anthropic launched its Cowork tool in January.
It remains unclear whether investments in AI will be profitable for companies that adopt the technology. Anthropic is betting that its new model will allow it to replicate the massive success of its Claude Code software but for other types of office work.
“We think Opus 4.6 is going to be an inflection point for knowledge work in many ways,” Dianne Penn, head of product management for Research, said in an interview with CNN ahead of the announcement.
One of these methods concerns the way Opus 4.6 processes data. Anthropic says it has expanded Opus’ pop-up window, which is the amount of information a model can remember at once, from 200,000 tokens to one million. Tokens are a unit of measurement referring to how AI models understand text. The longer and more complex a query is, the more tokens it requires.
Giving Claude the ability to process more information at once should allow him to handle more complex tasks, like making sweeping changes to entire code bases, Penn said.
With Opus 4.6, Claude will also know better when it takes more time to think through a request – a technique called reasoning – and when she needs to respond quickly.
Anthropic also claims that the new model outperforms OpenAI’s GPT-5.2 model on a baseline assessing how AI handles knowledge work in fields such as finance and law.
The new PowerPoint integration, available in search preview, will allow users to create slides using Claude, with the AI able to read layouts and fonts to create slides that align with the desired business model. This was particularly difficult because unlike Excel, which is primarily data-driven, PowerPoint slides involve making judgments about design elements such as colors and text placement, Penn said.
Anthropic claims that the files that Claude Opus 4.6 works on, like documents, spreadsheets, and slides, will be closer to being “production ready” the first time, meaning they should require less human intervention.
Although Anthropic’s Opus 4.6 improves Claude in non-technical work, the update also includes some improvements for software engineers. Coding tasks can be distributed among teams of agents instead of giving a single agent individual tasks, mimicking the way a human engineering team would operate, Anthropic says.
The release comes after software stocks collapsed this week following the release of plugins for Anthropic’s Cowork tool last Friday. These plugins, which allow users to tailor Anthropic’s Cowork tool to specific industries such as law, finance, sales and marketing, have raised concerns that the technology could replace specialized financial research and analysis software.
An exchange-traded fund that tracks the software industry had its worst day since April on Tuesday, falling nearly 6%. Thomson Reuters (SORTING) fell 15.83% on Tuesday, and Legalzoom (L.Z.) fell by almost 20% on the same day.
At the same time, questions and concerns are growing about the role AI will play in the workplace, particularly among entry-level tech jobs. Employment of recent computer science and mathematics graduates has decreased by 8% since 2022, according to a report from Oxford Economy last year. A Google study of September found that 90% of tech workers use AI in their jobs.
These kinds of concerns are “something we face and think about with every release of Claude and every product we put out,” Penn said, pointing to initiatives like the Anthropic Economic Index report, which studies the impact of AI on work.
But Jacob Bourne, a technology analyst at eMarketer, said these kinds of AI tools are unlikely to reshape the job market just yet. Security concerns will likely prevent many large businesses from using these types of tools, as they often require granting access to one’s files and browsing activity. (Penn said Anthropic works with its customers to ensure its technology meets their security and IT requirements.)
“The panic about this is probably misplaced,” Bourne said. “But I think it means legacy enterprise software vendors will have to continue to evolve.”