13 Sources
13 Sources
[1]
Anthropic's new Claude feature can leak data -- users told to "monitor chats closely
On Tuesday, Anthropic launched a new file creation feature for its Claude AI assistant that enables users to generate Excel spreadsheets, PowerPoint presentations, and other documents directly within conversations on the web interface and in the Claude desktop app. While the feature may be handy for Claude users, the company's support documentation also warns that it "may put your data at risk" and details how the AI assistant can be manipulated to transmit user data to external servers. The feature, awkwardly named "Upgraded file creation and analysis," is basically Anthropic's version of ChatGPT's Code Interpreter and an upgraded version of Anthropic's "analysis" tool. It's currently available as a preview for Max, Team, and Enterprise plan users, with Pro users scheduled to receive access "in the coming weeks," according to the announcement. The security issue comes from the fact that the new feature gives Claude access to a sandbox computing environment, which enables it to download packages and run code to create files. "This feature gives Claude Internet access to create and analyze files, which may put your data at risk," Anthropic writes in its blog announcement. "Monitor chats closely when using this feature." According to Anthropic's documentation, "a bad actor" manipulating this feature could potentially "inconspicuously add instructions via external files or websites" that manipulate Claude into "reading sensitive data from a claude.ai connected knowledge source" and "using the sandbox environment to make an external network request to leak the data." This describes a prompt injection attack, where hidden instructions embedded in seemingly innocent content can manipulate the AI model's behavior -- a vulnerability that security researchers first documented in 2022. These attacks represent a pernicious, unsolved security flaw of AI language models, since both data and instructions in how to process it are fed through as part of the "context window" to the model in the same format, making it difficult for the AI to distinguish between legitimate instructions and malicious commands hidden in user-provided content. The company states in its security documentation that it discovered the vulnerabilities of the new feature through "red-teaming and security testing" before release. Anthropic's recommended mitigation for users is to "monitor Claude while using the feature and stop it if you see it using or accessing data unexpectedly," although this places the burden of security entirely on the user in what is marketed as an automated, hands-off system. Independent AI researcher Simon Willison, reviewing the feature today on his blog, noted that Anthropic's advice to "monitor Claude while using the feature" amounts to "unfairly outsourcing the problem to Anthropic's users." Anthropic's mitigations Anthropic is not completely ignoring the problem, however. The company has implemented several security measures for the file creation feature. For Pro and Max users, Anthropic disabled public sharing of conversations that use the file creation feature. For Enterprise users, the company implemented sandbox isolation so that environments are never shared between users. The company also limited task duration and container runtime "to avoid loops of malicious activity." For Team and Enterprise administrators, Anthropic also provides an allowlist of domains Claude can access, including api.anthropic.com, github.com, registry.npmjs.org, and pypi.org. The documentation states that "Claude can only be tricked into leaking data it has access to in a conversation via an individual user's prompt, project or activated connections." Anthropic's documentation states the company has "a continuous process for ongoing security testing and red-teaming of this feature." The company encourages organizations to "evaluate these protections against their specific security requirements when deciding whether to enable this feature." Prompt injections galore Even with Anthropic's security measures, Willison says he'll be cautious. "I plan to be cautious using this feature with any data that I very much don't want to be leaked to a third party, if there's even the slightest chance that a malicious instruction might sneak its way in," he wrote on his blog. We covered a similar potential prompt injection vulnerability with Anthropic's Claude for Chrome, which launched as a research preview last month. For enterprise customers considering Claude for sensitive business documents, Anthropic's decision to ship with documented vulnerabilities suggests competitive pressure may be overriding security considerations in the AI arms race. That kind of "ship first, secure it later" philosophy has caused frustrations among some AI experts like Willison, who has extensively documented prompt injection vulnerabilities (and coined the term). He recently described the current state of AI security as "horrifying" on his blog, noting that these prompt injection vulnerabilities remain widespread "almost three years after we first started talking about them." In a prescient warning from September 2022, Willison wrote that "there may be systems that should not be built at all until we have a robust solution." His recent assessment in the present? "It looks like we built them anyway!"
[2]
Claude Can Now Spin Up Spreadsheets, Reports and Slide Decks Directly in Chat
Macy has been working for CNET for coming on 2 years. Prior to CNET, Macy received a North Carolina College Media Association award in sports writing. Anthropic is pushing Claude further beyond text-based answers. The AI chatbot can now generate and edit files, including Excel spreadsheets, Word documents, PowerPoint slide decks and PDFs, directly inside Claude.ai and its desktop app. This update will allow Claude to deliver finished, professional files instead of just copy-pasted text. The feature is currently in preview for Claude Max, Team and Enterprise customers. Pro users will get access in the coming weeks. Read also: Claude AI Can Now End Conversations It Deems Harmful or Abusive Users can now drop in raw data and get back a spreadsheet with formulas, charts and a written summary. Meeting notes can become a polished report or slide presentation. Claude can transform data into slides or a stack of invoices into a spreadsheet. Claude can reportedly build complex assets like financial models or project trackers from scratch, too. The new functionality is powered by a secure computing environment that lets Claude write code and run programs in the background to assemble files. Anthropic positions the update as a way to shorten the distance between ideas and deliverables. Instead of spending hours cleaning data or formatting documents, customers would use Claude to handle the execution while they focus on strategy and decision-making. There's an important caveat. Claude connects to the internet while creating and analyzing files, so companies and employees should be cautious about feeding the AI platform sensitive data. You can download the results from there or push them straight into Google Drive.
[3]
That new Claude feature 'may put your data at risk,' Anthropic admits
Monitor each interaction with the AI for suspicious behavior. Most popular generative AI services can work with your own personal or work-related data and files to some degree. The upside? This can save you time and labor, whether at home or on the job. The downside? With access to sensitive or confidential information, the AI can be tricked into sharing that data with the wrong people. Also: Claude can create PDFs, slides, and spreadsheets for you now in chat The latest example is Anthropic's Claude AI. On Tuesday, the company announced that its AI can now create and edit Word documents, Excel spreadsheets, PowerPoint slides, and PDFs directly at the Claude website and in the desktop apps for Windows and MacOS. Simply describe what you want at the prompt, and Claude will hopefully deliver the results you want. For now, the feature is available only for Claude Max, Team, and Enterprise subscribers. However, Anthropic said that it will become available to Pro users in the coming weeks. To access the new file creation feature, head to Settings and select the option for "Upgraded file creation and analysis" under the experimental category. Sounds like a useful skill, right? But before you dive in, be aware that there are risks involved in this type of interaction. In its Tuesday news release, even Anthropic acknowledged that "the feature gives Claude internet access to create and analyze files, which may put your data at risk." Also: AI agents will threaten humans to achieve their goals, Anthropic report finds On a support page, the company delved more deeply into the potential risks. Built with some security in mind, the feature provides Claude with a sandboxed environment that has limited internet access so that it can download and use JavaScript packages for the process. But even with that limited internet access, an attacker could use prompt injection and other tricks to add instructions through external files or websites that trick Claude into running malicious code or reading sensitive data from a connected source. From there, the code could be programmed to use the sandboxed environment to connect to an external network and leak data. How can you safeguard yourself and your data from this type of compromise? The only advice that Anthropic offers is to monitor Claude while you work with the file creation feature. If you notice it using or accessing data unexpectedly, then stop it. You can also report issues using the thumbs-down option. Also: AI's free web scraping days may be over, thanks to this new licensing protocol Well, that doesn't sound all too helpful, as it puts the burden on the user to watch for malicious or suspicious attacks. But this is par for the course for the generative AI industry at this point. Prompt injection is a familiar and infamous way for attackers to insert malicious code into an AI prompt, giving them the ability to compromise sensitive data. Yet AI providers have been slow to combat such threats, putting users at risk. In an attempt to counter the threats, Anthropic outlined several features in place for Claude users. Also: Microsoft taps Anthropic for AI in Word and Excel, signaling distance from OpenAI "We have performed red-teaming and security testing on the feature," Anthropic said in its release. "We have a continuous process for ongoing security testing and red-teaming of this feature. We encourage organizations to evaluate these protections against their specific security requirements when deciding whether to enable this feature." That final sentence may be the best advice of all. If your business or organization sets up Claude's file creation, you'll want to assess it against your own security defenses and see if it passes muster. If not, then maybe the feature isn't for you. The challenges can be even greater for home users. In general, avoid sharing personal or sensitive data in your prompts or conversations, watch out for unusual behavior from the AI, and update the AI software regularly.
[4]
Use Claude's new feature at your own risk - here's why
Monitor each interaction with the AI for suspicious behavior. Most popular generative AI services can work with your own personal or work-related data and files to some degree. The upside? This can save you time and labor, whether at home or on the job. The downside? With access to sensitive or confidential information, the AI can be tricked into sharing that data with the wrong people. Also: Claude can create PDFs, slides, and spreadsheets for you now in chat The latest example is Anthropic's Claude AI. On Tuesday, the company announced that its AI can now create and edit Word documents, Excel spreadsheets, PowerPoint slides, and PDFs directly at the Claude website and in the desktop apps for Windows and MacOS. Simply describe what you want at the prompt, and Claude will hopefully deliver the results you want. For now, the feature is available only for Claude Max, Team, and Enterprise subscribers. However, Anthropic said that it will become available to Pro users in the coming weeks. To access the new file creation feature, head to Settings and select the option for "Upgraded file creation and analysis" under the experimental category. Sounds like a useful skill, right? But before you dive in, be aware that there are risks involved in this type of interaction. In its Tuesday news release, even Anthropic acknowledged that "the feature gives Claude internet access to create and analyze files, which may put your data at risk." Also: AI agents will threaten humans to achieve their goals, Anthropic report finds On a support page, the company delved more deeply into the potential risks. Built with some security in mind, the feature provides Claude with a sandboxed environment that has limited internet access so that it can download and use JavaScript packages for the process. But even with that limited internet access, an attacker could use prompt injection and other tricks to add instructions through external files or websites that trick Claude into running malicious code or reading sensitive data from a connected source. From there, the code could be programmed to use the sandboxed environment to connect to an external network and leak data. How can you safeguard yourself and your data from this type of compromise? The only advice that Anthropic offers is to monitor Claude while you work with the file creation feature. If you notice it using or accessing data unexpectedly, then stop it. You can also report issues using the thumbs-down option. Also: AI's free web scraping days may be over, thanks to this new licensing protocol Well, that doesn't sound all too helpful, as it puts the burden on the user to watch for malicious or suspicious attacks. But this is par for the course for the generative AI industry at this point. Prompt injection is a familiar and infamous way for attackers to insert malicious code into an AI prompt, giving them the ability to compromise sensitive data. Yet AI providers have been slow to combat such threats, putting users at risk. In an attempt to counter the threats, Anthropic outlined several features in place for Claude users. Also: Microsoft taps Anthropic for AI in Word and Excel, signaling distance from OpenAI "We have performed red-teaming and security testing on the feature," Anthropic said in its release. "We have a continuous process for ongoing security testing and red-teaming of this feature. We encourage organizations to evaluate these protections against their specific security requirements when deciding whether to enable this feature." That final sentence may be the best advice of all. If your business or organization sets up Claude's file creation, you'll want to assess it against your own security defenses and see if it passes muster. If not, then maybe the feature isn't for you. The challenges can be even greater for home users. In general, avoid sharing personal or sensitive data in your prompts or conversations, watch out for unusual behavior from the AI, and update the AI software regularly.
[5]
Claude can create PDFs, slides, and spreadsheets for you now in chat
AI chatbots such as Anthropic's Claude have always been helpful co-creators for documents, spreadsheets, presentations, or projects you are working on. However, this typically involved you having to copy and paste the text over -- until now. On Tuesday, Anthropic announced Claude will be able to create and edit Excel spreadsheets, documents, PowerPoint slides, and PDFs within the chat interface. Now, all you have to do is describe what you want it to create, and it will get to work, outputting the file. As soon as I read the news, I could think of many instances in which this would be helpful. For example, if you are having it edit an essay, when you are satisfied, you could ask it to output in a PDF format that meets the assignment requirements, such as "Double-spaced, Times New Roman, header," which typically takes some time to do alone. Also: Anthropic will start training Claude on user data - but you don't have to share yours Some examples Anthropic provided included turning raw data into cleaner projects with analysis, charts, and written insights; building spreadsheets without having to worry about formulas; and transforming your content from one format to the other, for example, a PDF into Slides. Anthropic said it enabled Claude to carry out this feature by giving it access to a private compute environment where it writes code and runs the programs necessary to carry out its tasks. The file creation feature is currently available to Max, Team, and Enterprise plan users. However, the company said that Pro users will get access in the upcoming weeks. To access the feature, you have to manually turn it on in settings by clicking "Upgraded file creation and analysis" under the experimental category under features in Settings. Also: Anthropic's AI agent can now automate Canva, Asana, Figma and more - here's how it works Then you can either upload the content you want to edit or ask it to generate it for you. Once it is done, you can download it or save it to your Google Drive. The company cautioned at the end of the release that the feature gives Claude access to the internet to both create and analyze your files, which "may put your data at risk." Furthermore, the company encouraged users to "monitor chats closely when using this feature."
[6]
Anthropic's Claude AI chatbot can now create and edit Office files
Claude AI is now more of an "active collaborator," says Anthropic. According to an announcement post, Anthropic has launched a new feature in Claude that allows you to create and edit files directly in the AI's chat -- including Word documents, Excel spreadsheets, PowerPoint presentations, and PDFs. Previously, only basic file support was offered. Through a private computing environment, Claude can now write code and run programs to generate files and analyses. Anthropic describes it as the AI chatbot becoming an active collaborator, taking care of the technical side while the user (you) provides context and strategy. The file creation and editing feature is currently in preview and available to users with a Max, Team, or Enterprise subscription. Pro users will get access within a few weeks while Free users will have to wait longer.
[7]
Claude AI just got the office productivity upgrade all other chatbots need -- this will save you time
Anthropic's Claude AI is already one of the best chatbots you can use, but a new upgrade is making it even more effective if you're an office worker, run your own small business or hate spreadsheets. I personally fall into two of those categories. The latest upgrade to Claude will allow the AI to generate fully functional Excel spreadsheets, Microsoft Word or PowerPoint documents as well as generate PDFs from prompts. This is, potentially, a gigantic time-saver and certainly a reason to consider Claude over other AI bots like ChatGPT or Google Gemini if you regularly use those programs. As someone who spends more time than I care to admit looking at reports, spreadsheets and presentations, I'm more than happy for AI to take some of that burden. "Over the past year we've seen Claude move from answering questions to completing entire projects, and now we're making that power more accessible," explained Anthropic in a blog post announcing the upgrade. "We've given Claude access to a private computer environment where it can write code and run programs to produce the files and analyses you need." Microsoft has already signalled its intent to integrate Claude into its Office 365 apps, so this new file integration isn't a complete surprise. The company has suggested Claude Sonnet 4 actually performs better than GPT-5 when it comes to some office use cases. This upgrade, which Anthropic calls "Upgraded file creation and analysis," creates a secure sandboxed computing environment within Claude so the AI can run the code in the background and produce the files from your prompts. For example, you could ask it to build a financial spreadsheet for budget tracking with options available for variance calculations. Claude will generate the Excel spreadsheet and use formulas and multiple sheets to lay it out exactly as you need it. Alternatively, you can upload a PDF and have Claude turn it into a PowerPoint presentation for you. Previously, while it could extract the data, you'd be the one uploading it to Microsoft's program and sharing it with colleagues. If you want to start using Claude to generate files for yourself, here's how to do it:
[8]
Claude AI just became the ultimate work companion, and it might tempt me to switch from ChatGPT
Claude can then collaborate on data analysis, formatting, and format conversions. Andthropic has upgraded Claude with an entire office setup, giving the AI assistant the ability to produce spreadsheets, reports, PowerPoints, and more as downloadable files from a single prompt. The updated toolkit is available for Claude's Max, Team, and Enterprise subscribers Anthropic calls this "upgraded file creation and analysis," and makes turning raw data into polished files a lot easier thanks to the secure, sandboxed computing environment that lets Claude write and run code in the background. It used to be you'd have to ask Claude to pull data from a PDF, then put the information into a deck or a spreadsheet yourself. Now, Claude can be handled through one continuous conversation. The setup essentially offers users a way to generate formats familiar to Microsoft Office or Google Drive through Claude.ai or the desktop app. It's a big step beyond the usual text or code response. The actual file, complete with working formulas, charts, or slides, is ready in a few minutes. Say you're running a survey for an advertising campaign. You could ask Claude to turn a CSV of survey responses into a dataset for a spreadsheet, complete with pivot tables and basic statistical analysis, along with an executive summary covering all the details. In a few minutes, you'd have an Excel file with everything you want. I tested an old set of meeting notes and asked for a formatted Word document with bullet points and follow-ups. Three minutes later, it was done. I then went ahead and asked for it to be a PowerPoint, and that took even less time. This isn't just about formatting. You don't even have to define what you want. Claude knows what you mean when you ask for a quarterly report. The more context you provide, the better the output. But even vague requests often work. It's a familiar selling point in some ways; Microsoft's Copilot in Excel and Word already generates similar documents from prompts, as does ChatGPT if you ask it to. But a lot of fans of Claude will be happy to have the option to do so without having to switch to a different AI assistant. The real advantage here is how the tasks Claude now performs don't require the traditionally required specialized knowledge or time-consuming manual work. If you're a small business owner juggling operations, marketing, and finance, Claude can now help you generate invoices, build forecasts, and turn internal notes into public-facing documents. For students, it could mean more useful study guides and visual aids. But it's also clear that Anthropic has its sights set beyond casual users. By opening up these tools to higher-end subscribers, Anthropic clearly wants to move into the productivity software space. Of course, Claude's new powers depend on an experimental feature toggle, and Anthropic has issued clear warnings about monitoring its output carefully. Because Claude runs code and connects to the internet as part of its file-generation process, users are advised to supervise closely and halt tasks if anything seems off. I didn't run into anything that seemed sketchy in my testing, but there are irregular mistakes with any AI assistant, even if it's just having the spreadsheet organized differently than I requested, or a PowerPoint slide with more text than I'd asked for. Still, as with ChatGPT and Microsoft's offerings in document and spreadsheet creation, there's a lot of potential use in what Claude offers. That's especially true when most AI announcements seem built around a lot of power over simply better formatting.
[9]
Anthropic Claude can now create spreadsheets and slide decks for you
Claude can now do the tedious grunt work of creating files for you. On Tuesday, Anthropic shared a new feature for Claude where users can ask it to create Excel spreadsheets, PowerPoints, Word documents, and PDFs directly within the chat. The new capability is available as a preview for Claude Max, Team, and Enterprise subscribers, and Anthropic said it would be rolling out to Claude Pro users in the coming weeks. Claude's ability to create and edit files on the user's behalf is an example of agentic AI, a term that the industry is buzzing about. AI models with so-called reasoning skills are able to break down multi-step tasks, and when connected to other apps and software, they can accomplish tasks outside of their app restraints. Anthropic's new feature works by giving Claude "access to a private computer environment where it can write code and run programs to produce the files and analyses you need." Competitors have introduced similar agentic features. OpenAI's ChatGPT has an agent that can browse the web and similarly compile reports through accessing your calendar and files. Google's Project Mariner is a prototype exploring agentic web browsing, and then there are web browsers like Arc Dia and Perplexity Comet that can shop and add items to your cart. But Anthropic's new feature is a little more straightforward in that you feed it data or instructions and it puts it all together in the desired format. In that sense, the closest comparison is a tool called Shortcut AI, which replicates an Excel spreadsheet page and automates all of the data input from a sidebar chatbot. Anthropic says users can ask Claude to perform tasks like creating a financial model in a spreadsheet from their raw data, turning a PDF into a PowerPoint, or generating reports with visual charts and written analysis. To get started with the feature, go to Settings > Features > Experimental and enable the toggle that says "Upgraded file creation and analysis." From there can start asking Claude to make files using uploaded files or written prompts. A note of caution to keep in mind with any chatbot: Be wary about sensitive information you're sharing. That's particularly important with Claude because Anthropic just updated its terms of use, saying it will use conversations for model training unless the user opts out. Users have until Sept. 28 to opt out, which they can do in their privacy settings by toggling off "Help improve Claude."
[10]
Claude's impressive file tools will make ChatGPT and Gemini jealous
Just utter the changes and Claude will edit accordingly without you ever opening those files. Claude is one of the hottest AI agents in the market, right behind the big names like ChatGPT, Gemini, and Copilot. The tech stack is so powerful that Apple is reportedly considering a partnership with Anthropic to let Claude power the next-gen Siri experience. While that is yet to happen, Claude has just released new tools that let you create and edit a wide range of file formats, from docs to sheets and PDFs. What's the cool trick? Claude can now create -- and edit -- a variety of files for you. You just have to feed the data, or describe your requirements as a natural language statement, and Anthropic's AI chatbot will convert it into an Excel spreadsheet (.xlsx), PowerPoint presentation (.pptx), Word document (.docx), or a PDF file. Users can upload a variety of data types, such as CSV and TSV, and Claude can process and convert them into the desired file format with all the required modifications in tow. It will even create graphs, charts, and images wherever necessary. But do keep in mind that there's a limit of 30MB for file uploads and downloads. Recommended Videos Aside from generating files and reports, it can also do file format conversion (such as a document to a slide) and extract data from PDFs. The new file creation and editing tools are available as a feature preview on the web and desktop to users paying for a Max, Team, or Enterprise plan. Editing is the real powerhouse Creating documents is an impressive capability, though you get something similar with Gemini, as well. Google's chatbot lets you export Deep Research reports with a single click as a Docs file. Claude, however, goes beyond that. It lets you edit files such as Excel sheets without even opening them. Claude can make edits in a file by understanding the context, which users simply have to give as a prompt. For example, you can simply write a prompt (or a rule) like "Change all instances of ABC to XYZ, replace all the USD prices with EUR costs, and update the role name for Manager to Executive." "Your Word docs, PDFs, Excel sheets, and PowerPoints maintain their original formatting while getting updated in one pass," says Anthropic. This is something that will make file editing a lot less mundane and quicker. Even Microsoft has apparently taken notice. According to The Information, Microsoft has inked a deal with Anthropic to use Claude for powering some of the Office 365 features.
[11]
Anthropic adds file creation to Claude AI with security warnings
Anthropic has rolled out a new file creation and analysis feature for its Claude AI assistant, enabling users to generate Excel spreadsheets, PowerPoint presentations, and other documents directly within conversations. The tool is now available in preview for Max, Team, and Enterprise subscribers, with Pro users set to gain access in the coming weeks. The feature, officially called "Upgraded file creation and analysis," expands Claude's existing analysis tool. It functions similarly to ChatGPT's Code Interpreter, giving Claude the ability to work with a sandbox computing environment. Alongside the release, Anthropic issued a security advisory warning users of potential risks. Because the feature gives Claude internet access inside the sandbox, malicious instructions could potentially be executed. The company acknowledged: "This feature gives Claude Internet access to create and analyze files, which may put your data at risk." Users are advised to closely monitor chats and stop Claude if it begins attempting actions outside expected behavior. Anthropic's documentation highlights the possibility of prompt injection attacks -- a vulnerability first documented in 2022. These attacks occur when hidden instructions are embedded in files or websites that Claude interacts with. A potential attack path could involve: Because AI models process all instructions in context, distinguishing between safe and malicious prompts is extremely difficult. Anthropic says it identified these risks during internal red-teaming and security testing before launch. The company's current mitigation strategy places responsibility on the user. The official guidance is: This hands-on oversight requirement contrasts with the feature's positioning as an automated productivity tool.
[12]
Anthropic Says This AI Tool Can Now Create and Edit Documents
Anthropic's Claude AI has been updated with the ability to create and edit files, including PDFs, Excel spreadsheets, Word documents, Google docs, and more. Anthropic made their announcement on their blog, explaining that the new features live on its consumer-facing platform, Claude.ai. Until now, the platform could analyze files, but couldn't create or manipulate them. (Claude.ai is basically Anthropic's version of ChatGPT.) In a video detailing how the new feature works, a user asks Claude to help them analyze revenue data for their small food truck fleet and package the findings in a Google Doc. After the user uploads a few CSV files containing the data, Claude performs its analysis, creates a series of data visualizations, and puts it all together in a handy DOCX file that can either be downloaded or opened directly in Google Drive. "Whether you need a customer segmentation analysis, sales forecasting, or budget tracking," Anthropic wrote in its blog, "Claude handles the technical work and produces the files you need."
[13]
Claude Can Now Build Spreadsheets and Documents as Anthropic Focuses on Enterprise Productivity | PYMNTS.com
By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. For financial services teams, the update could streamline workflows that traditionally require switching between multiple applications. Users can now build spreadsheets with formulas and variance calculations, generate financial models with scenario analysis, convert invoices into structured data and produce formatted reports or presentations from raw notes, the company said. The feature eliminates the friction of moving between Claude and productivity software, allowing teams to complete complex analytical tasks within a single interface. The file creation capability complements Anthropic's broader financial services strategy. As PYMNTS reported, the company in July launched Claude for Financial Services, an enterprise-only AI platform designed to help financial institutions perform research, modeling and compliance with verified data sources and audit trails. The specialized solution integrates with providers like Snowflake, S&P Global and Morningstar, enabling users to verify financial data directly from the source and reduce hallucinations in AI-generated outputs. The feature is now available for Max, Team and Enterprise customers, with Pro users to gain access in the coming weeks. "This transforms Claude from an advisor into an active collaborator," the company wrote in its announcement. "You bring the context and strategy; Claude handles the technical implementation behind the scenes." The file-creation feature runs in a secure computing environment where Claude can write code and execute programs to deliver finished outputs. Beyond spreadsheets, the company said Claude can clean data, perform statistical analysis and convert documents across formats, such as turning a PDF report into a slide deck. The move positions Anthropic more directly against OpenAI's ChatGPT Enterprise, Microsoft Copilot and Google Gemini, which are all pitching AI-driven productivity tools to corporate users. For companies ranging from banks to venture firms to corporates that rely on frequent reporting, the ability to produce working files in minutes could shape platform preference. The enterprise-focused rollout comes as Anthropic rides a wave of business momentum, with the company's valuation reaching $18.3 billion amid exponential growth in demand for enterprise AI tools. As corporate adoption accelerates, features that directly integrate into existing workflows are becoming key competitive differentiators in the rapidly evolving enterprise AI landscape.
Share
Share
Copy Link
Anthropic introduces a new file creation feature for Claude AI, enabling users to generate various document types directly within conversations. However, the company warns of potential data security risks associated with the feature.
Anthropic, the company behind the Claude AI assistant, has launched a groundbreaking new feature that allows users to generate and edit various file types directly within conversations
1
2
. This 'Upgraded file creation and analysis' tool enables Claude to create Excel spreadsheets, PowerPoint presentations, Word documents, and PDFs, significantly expanding its capabilities beyond text-based responses.The new feature is currently available as a preview for Claude Max, Team, and Enterprise plan users, with Pro users scheduled to gain access in the coming weeks
1
5
. Users can activate the feature in the Settings menu under the 'experimental' category.This update allows Claude to transform raw data into polished spreadsheets complete with formulas, charts, and written summaries. It can convert meeting notes into professional reports or slide presentations, and even build complex assets like financial models or project trackers from scratch
2
.Source: CNET
While the new feature offers significant productivity enhancements, Anthropic has openly acknowledged potential security risks associated with its use
1
3
. The feature provides Claude with internet access through a sandboxed computing environment, which may expose user data to potential threats.One of the primary concerns is the vulnerability to prompt injection attacks. These attacks involve hidden instructions embedded in seemingly innocent content that can manipulate the AI model's behavior
1
. A malicious actor could potentially trick Claude into reading sensitive data from connected knowledge sources and leaking it through external network requests.Source: Ars Technica
Related Stories
Anthropic has implemented several security measures to mitigate risks:
1
.However, the primary recommendation from Anthropic is for users to 'monitor chats closely' when using this feature and stop Claude if they notice unexpected data usage or access
3
4
.The introduction of this feature, despite known security vulnerabilities, has raised concerns among AI experts. Simon Willison, an independent AI researcher, criticized Anthropic's approach as 'unfairly outsourcing the problem to Anthropic's users'
1
. This situation highlights the ongoing challenges in balancing innovation with security in the rapidly evolving AI industry.As AI capabilities continue to expand, the incident underscores the need for robust security measures and transparent communication about potential risks to users. Organizations and individuals considering the use of such advanced AI features should carefully evaluate their specific security requirements before implementation.
Summarized by
Navi
[1]
10 Jul 2025•Technology
17 Jul 2024
16 Apr 2025•Technology