December 2025 Summaries

Agentic process automation is an advanced AI-driven approach that allows systems to autonomously manage complex business workflows by understanding context, reasoning over prior steps, and executing tasks without relying on fixed rules or manual handoffs. This method enhances business efficiency by enabling faster, more consistent decision-making and reducing repetitive tasks, thus allowing teams to focus on strategic work. By interpreting documents and identifying next steps autonomously, agentic AI systems support seamless process progression and better exception handling, with Gartner predicting that by 2028, autonomous decisions will constitute at least 15% of daily work decisions. Companies can choose suitable agentic tools by considering capabilities such as permission-aware content access, native document understanding, and strong governance guardrails, with Box providing a platform for secure, end-to-end agentic automation. Unlike traditional or robotic process automation, which relies on predefined rules, agentic AI uses reasoning and real-time information to adapt to variable inputs and changing conditions, making it ideal for dynamic, decision-heavy workflows.

Vornado Realty Trust has undergone a significant transformation by integrating AI into its operations, revolutionizing how the company manages its vast data resources. Under the leadership of Robert Entin, the company transitioned its content to the cloud nearly a decade ago, setting the stage for leveraging AI to handle complex lease documents, invoices, and other unstructured data. By utilizing AI-powered solutions like Box Hubs, Vornado has enhanced its efficiency in processing loan documentation, earnings call transcripts, and financial analysis, allowing for quicker and more accurate information retrieval. This transformation not only streamlines operations but also offers a competitive edge in the commercial real estate sector, as Vornado continues to innovate by combining AI capabilities with robust content management strategies. Through hands-on experimentation and a strategic approach to data readiness, Vornado provides a model for other organizations facing similar challenges in data management and operational efficiency.

Artificial intelligence (AI) is revolutionizing knowledge management by enhancing how organizations access, interpret, and utilize information. Traditional systems often struggle with manual tagging, outdated files, and disconnected systems, while AI efficiently organizes and retrieves content by understanding context and linking related ideas. AI-powered knowledge management automates metadata enrichment, improves search accuracy, and reveals connections across disparate files, thereby reducing manual efforts and supporting faster decision-making. With tools like Box AI, companies can implement secure semantic search, automated workflows, and intelligent content management, resulting in faster information retrieval, higher data quality, and consistent knowledge access. However, maintaining accurate and updated repositories is crucial to mitigate risks like inaccurate AI outputs. AI systems that integrate seamlessly with existing workflows, understand intent, and provide personalized insights are considered the most effective in enhancing enterprise content strategy.

In 2025, the Box Developer Relations team significantly advanced the integration of AI into enterprise content management by publishing 57 pieces of content, including technical guides on AI integrations with Box AI API and emerging AI frameworks like LangChain, Pinecone, and OpenAI. They embraced the Model Context Protocol (MCP) ecosystem, assisting developers in building AI agents that integrate Box’s Intelligent Content Management with advanced AI tools. Additionally, they organized a series of events, including the flagship BoxWorks in San Francisco, to explore the future of enterprise AI and foster a global developer community through hands-on education and strategic partnerships. The team also produced 25 video tutorials to showcase AI-powered development, featuring advanced AI models and practical applications with Box’s platform. Key team members like Scott Hurrey, Andrew Wong, Rui Barbosa, and Olga Stefaniuk contributed to these efforts through content creation, event participation, and community engagement, enhancing Box’s position as a leader in Intelligent Content Management while encouraging developers to apply new insights and innovations.

American Humane Society, a leading organization in animal welfare, is leveraging artificial intelligence (AI) to enhance its mission of protecting over 1.5 billion animals worldwide. Under the leadership of Chief Technology Officer Karthik Devarajan, AI is being used to detect patterns in extensive data sets, such as temperature logs and feeding schedules, which are otherwise too complex for manual analysis. By integrating Box AI, the organization has increased efficiency by 90% in certain processes, allowing staff to focus more on direct animal care. The approach to AI is methodical, starting with small, targeted solutions that build trust and demonstrate value before expanding to broader applications. This strategy aligns with their long-standing commitment to compassion and mission alignment, ensuring that AI acts as a force multiplier for animal protection rather than a disruptive force. Despite challenges, such as managing large datasets and ensuring data governance, the organization has successfully integrated AI into its workflows, saving time and allowing staff to focus on their core mission of animal welfare. This transformation reflects American Humane Society's philosophy of using technology to amplify kindness and compassion, positioning AI as a natural part of their operations rather than a standalone project.

AI-powered enterprise search tools are increasingly being adopted by large organizations to address challenges in locating work-related information scattered across multiple platforms and repositories. These tools leverage advanced models to understand user intent and context, offering precise information retrieval while reducing time spent searching and minimizing errors from outdated or siloed content. By moving beyond traditional keyword-based searches, AI systems deliver accurate, permission-aware results, aiding in faster decision-making and improving productivity. The enterprise search market is projected to grow significantly due to the surge in unstructured data and the demand for efficient information retrieval. Companies like Box provide solutions that integrate semantic and intent-driven search capabilities, ensuring results are aligned with access levels and compliance requirements. These AI tools also enable the summarization of documents, extraction of key insights, and identification of related files, ultimately enhancing knowledge discovery and governance within enterprises.

AI metadata significantly enhances the management and utilization of unstructured content by automating the extraction and tagging processes, which traditionally relied on manual effort and were prone to inconsistencies and inefficiencies. Advanced AI models can interpret complex documents by identifying and connecting key entities, clauses, and other details, thus improving search accuracy, governance, and automation across enterprises. This capability is crucial as most valuable business insights are often hidden in unstructured data that traditional systems cannot fully process. By employing AI-driven metadata extraction, organizations can achieve higher returns on investment through streamlined data management, improved decision-making, and reduced operational friction. The adoption of consistent metadata practices, centralized repositories, and standardized extraction schemas further strengthens AI-driven workflows, while integration into governance frameworks ensures compliance and risk management. As the demand for efficient content management grows, AI metadata solutions are becoming increasingly vital to convert everyday files into structured, actionable information.

In 2025, enterprise AI matured significantly, transitioning from experimental stages to becoming an integral part of business operations. This evolution was marked by five pivotal shifts: moving from prompt engineering to agent engineering, where comprehensive AI environments were designed; the transformation of AI from assistive tools to autonomous agents capable of managing complex workflows independently; a foundational approach to security that integrated permissions-aware architectures from the start; the adoption of specialized models that replaced one-size-fits-all solutions, allowing for strategic task routing; and the critical importance of trust, with organizations focusing on governance, observability, and auditability to ensure transparency and accountability in AI deployments. These changes highlight the growing emphasis on building intelligent, secure, and human-centric AI systems that generate real business value.

The Box AI Professional certification is a free, self-paced program designed to equip professionals with the skills needed to apply generative AI to business workflows using Box. It offers a blend of practical lessons, hands-on exercises, and a final exam, covering topics like AI fundamentals, Box AI capabilities, and responsible AI adoption in enterprise settings. Participants can expect to gain practical, job-ready skills, enabling them to identify high-value AI use cases, integrate AI into their workflows, and scale solutions across teams. The program includes six on-demand courses totaling about 3.5 hours and culminates in a 45-minute exam to earn a shareable LinkedIn certification, demonstrating AI proficiency and enhancing productivity and decision-making within organizations.

Box, an AI-first company, is transforming its operations through an ideation phase that fosters the development of impactful AI solutions, known as "hero agents," which are integrated into business processes to drive strategic growth. The ideation process at Box involves examining internal inefficiencies and exploring potential AI applications through organic experimentation and strategic prioritization, often facilitated by hackathons and submission forms. A significant example is the Intelligent Prospecting Agent, designed to streamline sales outreach by reducing the time sales reps spend identifying target accounts and crafting messages, effectively turning it into a critical business asset. Box emphasizes a bottom-up approach, allowing employees from any department to pitch ideas, while functional leaders play a key role in prioritizing and championing these innovations to ensure alignment with broader business objectives. This iterative feedback loop not only encourages innovation but also helps to create a scalable framework for AI integration across various business functions, with ongoing pilot programs testing the practical impact of these AI-driven solutions.

By 2025, enterprise AI has evolved from basic chatbots to advanced agentic systems capable of complex tasks such as extracting insights from unstructured data, structuring metadata, and enhancing security measures. Box CEO Aaron Levie envisions AI agents becoming widely available, transforming capabilities that were once rare. In 2026, strategic "hero agents" optimized for specific functions will offer more value than numerous underutilized ones, while developers' roles will shift from coders to orchestrators, integrating AI deeply into work processes across all departments. As AI becomes integral to operations, cybersecurity must adapt to new challenges, emphasizing strong governance and digital hygiene. Organizational design will play a crucial role in overcoming human bottlenecks, allowing employees to focus on strategic tasks and fostering collaboration between AI and human workers. Japan is highlighted as a leader in AI-human collaboration due to its labor shortage, demonstrating how AI can enhance workforce capacity and unlock business value.

Gemini 3 Flash, the latest release in enterprise AI, demonstrates significant improvements over its predecessor, Gemini 2.5 Flash, particularly in precision extraction and contextual understanding of complex documents. The model exhibits a 10 percentage-point increase in accuracy and excels in extracting critical data points from challenging layouts, unlocking new categories of data for automated extraction. It shows notable advancements in handling long documents, with a 6-point accuracy lift, and excels in extracting extensive metadata from numerous fields in a single file, achieving a 13 percentage-point improvement. Gemini 3 Flash also enhances data extraction from PDFs and images, with 10 and 9-point accuracy increases, respectively, making it highly effective for business operations that rely on these formats. These advancements enable enterprises to automate complex workflows in wealth management, insurance, legal, and healthcare sectors by accurately extracting data from dense, mixed-media files, thus shifting workflows from manual entry to automated processes with high confidence.

The rapidly advancing AI-first era is revolutionizing how businesses handle vast amounts of unstructured data by leveraging AI to extract and organize metadata, thus driving automation and efficiency across various departments. Companies like Box are at the forefront, using AI Extract Agents to convert unstructured files into structured, machine-readable intelligence, enabling smarter business decisions and streamlined content discovery. These agents, such as the Box Standard and Enhanced Extract Agents, cater to different document complexities and are instrumental in transforming static documents into dynamic entities that integrate with business systems like Salesforce. This technological shift not only automates data extraction but also enhances operational efficiency, reduces manual efforts, and saves costs significantly, as demonstrated by Box's own use cases. The continual evolution of AI models, like the transition to Gemini 2.5 Pro, ensures improved accuracy in data processing, positioning Box as a leader in intelligent content management. As AI innovations continue to emerge, businesses are poised to unlock the full potential of their unstructured data, paving the way for a new era of enterprise technology characterized by smarter decisions, efficient workflows, and heightened productivity.

In the AI-first era, smaller businesses are leveraging their agility to achieve faster financial returns from AI investments compared to larger enterprises, according to a report by the Wharton School and GBK Collective. Companies with less than $2 billion in revenue are seeing significant gains because they can swiftly implement necessary workflow changes and avoid the experimentation phase that often bogs down larger corporations. Large enterprises, identified as Tier 1, face challenges in translating AI investments into profit due to organizational complexity and inertia, while mid-sized and smaller firms report quicker ROI realization. The report highlights the importance of redefining business workflows in an AI-first manner to maximize AI benefits, emphasizing that a complete overhaul of business processes, rather than incremental integration, is crucial. Employee training is also vital, as skill atrophy can hinder AI scaling, and company success in AI adoption is linked to how well teams are acquainted with the technology. Aaron Levie, Box CEO, underscores the need for businesses, regardless of size, to reengineer workflows to harness AI effectively, while Olivia Nottebohm, Box COO, advocates for a strategic approach to AI transformation rather than ad hoc implementation.

Box has unveiled a new AI-first developer documentation site designed to enhance developer productivity and streamline the process of working with Box APIs. This beta version, built on the Mintlify platform, includes AI-powered chat and advanced search features, an interactive API playground, and an MCP server that integrates with IDEs for seamless access to Box technical content. The platform's features, such as dark mode, a simplified homepage, and improved navigation, are aimed at providing a more intuitive user experience. Users can interact with the AI assistant to ask questions, troubleshoot errors, and receive code examples, while the search functionality aggregates information from various Box resources. An interactive API testing experience allows developers to experiment with Box APIs directly within the documentation. As further updates are planned, Box invites users to explore the new portal and provide feedback through their developer forum.

The biopharmaceutical industry is enhancing patient outcomes through innovations in drug therapies and medical devices, facing pressures to accelerate launches while complying with regulations like 21 CFR Part 11 and HIPAA. Challenges arise from outdated content management systems that slow down research and development (R&D) processes. Intelligent Content Management platforms, such as Box, offer a solution by centralizing content in a secure, AI-ready environment, enabling seamless collaboration and faster drug development. Box helps streamline operations by integrating post-merger content, managing e-signature workflows, and migrating records to the cloud, which reduces silos and archives regulated data. By harnessing AI, Box provides actionable insights and fosters collaboration, ensuring transparent data exchange and audit-readiness. These capabilities allow biopharmaceutical companies to reduce risks and bring therapies to market more efficiently, transforming R&D processes and unlocking the value of their content.

Document lifecycle management (DLM) is a structured approach that governs the creation, review, storage, sharing, and disposal of documents, aiming to enhance efficiency, compliance, and risk reduction across organizations. Intelligent Content Management, powered by AI, elevates DLM by automating processes, improving search capabilities, and ensuring compliance through data encryption and role-based access controls, thus minimizing manual efforts and enhancing productivity. The integration of DLM software like Box with existing business tools simplifies workflow management, accelerates collaboration, and provides complete visibility and control over content ecosystems, contributing to significant productivity savings and streamlined operations. By applying best practices such as defining data retention policies, maintaining version control, and automating workflows, organizations can protect sensitive data, ensure compliance with standards like ISO 9001 and ISO 27001, and optimize their document management strategies. The choice between cloud-based and on-premise DLM solutions depends on factors like scalability, accessibility, and compliance needs, with cloud solutions offering advantages in terms of scalability and reduced IT overhead, while on-premise solutions offer greater control and customization.

Box's approach to becoming an AI-first company involves a structured, four-phase deployment framework to integrate AI agents effectively into business workflows. The process begins with ideation, where internal inefficiencies are identified, and potential AI solutions, termed "big bets," are prioritized for their high repeatability and critical thinking potential. These ideas then move into a pilot phase, where selected agents are tested over 3-6 months to validate performance metrics and user experience. Successful pilots advance to a rollout preparation phase, ensuring agents are production-ready with reliable data pipelines, governance protocols, and clear documentation. The final phase, scaled adoption, focuses on change management to embed these agents into daily use, supported by comprehensive training and accountability measures. This iterative process is sustained by continuous feedback loops and regular reviews, enabling ongoing innovation and refinement of AI integration across the company.

OpenAI’s GPT-5.2 marks a significant advancement in AI models by enhancing speed and complex reasoning capabilities, becoming more specialized across various industries such as media, financial services, and healthcare. Evaluated using Box AI Enterprise Eval, it shows notable improvements in latency and reasoning scores, outperforming its predecessors, GPT-5 and GPT-5.1, particularly in handling tasks with long contexts and complex workflows. In media and entertainment, it quickly analyzes box office data to inform distribution strategies, while in financial services, it accelerates due diligence by efficiently comparing loan documents. In life sciences and healthcare, GPT-5.2 aids researchers by synthesizing extensive clinical data into actionable insights, demonstrating its ability to transform knowledge work across high-stakes environments. This new model is available for all Box AI customers, promising to handle both straightforward and highly specialized content with increased agility and accuracy.

Box has been chosen by the Heart Research Institute (HRI) to centralize and secure its research content, aiming to enhance AI-powered insights, integrate seamlessly into research applications, and maintain compliance with healthcare and scientific standards. HRI, known for its pioneering cardiovascular research, will utilize Box's Enterprise Advanced plan to streamline content management, accelerate researcher onboarding, and digitize recruitment workflows, all while leveraging secure AI agents for data analysis. This collaboration underscores Box's role in assisting organizations within regulated industries to maximize their content's potential, facilitating impactful innovations and improving health outcomes.

Enterprise AI agents are intelligent systems designed to streamline high-volume, content-intensive work within organizations by understanding context, retrieving accurate answers, and completing routine tasks, thereby enhancing productivity and reducing bottlenecks. Unlike basic assistants, these agents integrate deeply with business environments to autonomously orchestrate workflows, interpret content, and make context-sensitive decisions while adhering to security and data governance protocols. They improve efficiency by reducing review cycles, minimizing manual interventions, and allowing teams to focus on higher-value activities. Agentic AI systems, which differ from generative AI by focusing on autonomous decision-making and task coordination, enhance enterprise operations through content understanding, task planning, and cross-system execution. Successful deployment of these agents requires strong governance, deep system integrations, and scalable automation capabilities, with platforms like Box offering solutions that provide secure, content-aware automation to accelerate work and maintain data control. As enterprises increasingly adopt AI technologies, the focus is on achieving reliable, secure, and scalable automation to maintain a competitive edge and support complex workflow operations.

Box AI Agents are now available in ServiceNow's AI Agent Marketplace, offering enhanced data extraction and content insights within ServiceNow's NowAssist. These agents streamline tasks such as finding contract terms, summarizing documentation, and extracting RFP requirements by acting as a secure content layer with access control. The integration includes two main capabilities: the Box Data Extraction Agent and the Box Content Insights Agent. The Data Extraction Agent identifies and captures critical data from Box files, while the Content Insights Agent analyzes and summarizes content, providing concise takeaways. This collaboration between Box and ServiceNow allows for automated and intelligent content workflows, empowering organizations to build custom agents for secure content interaction. The integration is designed to enhance enterprise productivity by enabling faster, more confident decision-making and maximizing content value through natural conversation interactions with Now Assist. Box customers can now access these AI capabilities directly within ServiceNow's interface, facilitating efficient management of content-dependent tasks.

Higher education institutions are increasingly integrating artificial intelligence (AI) into their operations and educational environments to optimize costs, streamline processes, and enhance strategic decision-making. At EDUCAUSE 2025, a shift was observed from theoretical discussions to practical AI implementations across campuses, with a focus on creating a collaborative human-AI culture. The 2026 EDUCAUSE Top 10 priorities highlight the importance of empowering stakeholders to engage with AI safely and creatively, emphasizing initiatives like the "Human Edge of AI" and "Knowledge Management for Safer AI." Box is playing a pivotal role by offering tools for content-aware AI, AI governance controls, and intelligent content management to ensure secure and innovative environments. The company's offerings, such as Box AI Studio and Institutional Storage, support these priorities by facilitating custom AI solutions and centralized management of institutional data while maintaining robust security measures. By aligning with these priorities, Box enables institutions to leverage AI's transformative potential responsibly, fostering a data-centric culture and collaborative cybersecurity while safeguarding the essential human elements of education.

When Google's Agent Development Kit is integrated with Box's Intelligent Content Management platform, creating sophisticated AI agents becomes significantly streamlined, reducing what typically requires extensive API integration and custom tooling to just eight lines of code, thanks to the model context protocol (MCP) and the enhanced Box MCP server. This integration allows AI agents to access enterprise-grade content operations with ease, facilitating tasks such as document generation, metadata management, and complex business workflow orchestration. The enhanced MCP server now supports Server-Sent Events (SSE) for better error handling and deployment flexibility, enabling AI frameworks to communicate with external tools without custom integrations. A practical scenario includes insurance claim adjusters quickly generating preliminary reports from structured data extracted from documents, showcasing the efficiency of this integration. While the development edition offers a fast and simple setup for prototyping and concept validation, production deployments require considerations for authentication, multi-tenancy, and scalability. This integration represents a significant advancement in bridging AI capabilities with enterprise needs, allowing developers to prototype quickly and explore new possibilities in AI and enterprise content management.

Box, a company known for its extensive customer support knowledge base, successfully streamlined its customer service operations by implementing an AI-powered support portal, significantly enhancing response times and customer satisfaction. By integrating automated knowledge systems and leveraging AI to parse natural language queries, Box was able to reduce the time spent searching for information across various sources like Zendesk and Slack, thereby deflecting an estimated 50% of service questions and saving over 1,000 hours annually. This initiative allowed teams to focus on more complex customer interactions, improved first-contact resolution rates, and highlighted gaps in existing documentation to address. The integration preserved existing workflows, ensuring rapid adoption without disruption, and serves as a model for any organization looking to optimize support efficiency by centralizing access to critical information.

Data governance for artificial intelligence (AI) is crucial as AI systems become integrated into core business functions, necessitating traceable and compliant model decisions based on verified enterprise data. Effective AI data governance involves establishing frameworks and best practices that ensure AI systems access only accurate, compliant data, thereby minimizing risks and aligning with privacy and security regulations. The framework involves setting clear AI goals, assigning cross-functional ownership, classifying sensitive content, enforcing granular access and compliance policies, and continuously refining governance rules. The integration of tools like Box can help centralize security, compliance, and lifecycle controls, fostering responsible AI deployment while protecting sensitive information. These efforts are essential as organizations adapt their governance strategies to support AI and automation, ensuring the technology improves productivity without compromising security, trust, or regulatory compliance.

AI is becoming integral to various business functions, but the traditional approach of training or fine-tuning AI models may not be the most effective or feasible for most enterprises due to high costs, complexity, and security risks. Box CTO Ben Kus and AI expert Meena Ganesh advocate for retrieval augmented generation (RAG) combined with AI agents as a smarter and more secure alternative. This approach enables AI systems to use existing models to access relevant data securely, minimizing the need for ongoing model management and reducing the risk of data leakage. RAG allows enterprises to efficiently harness AI for context retrieval and workflow automation without the resource-intensive demands of model training or fine-tuning. While custom training might be necessary for certain edge cases, for the majority of businesses, leveraging AI agents through RAG is a more practical and scalable solution.

Box AI Metadata Extraction, integrated within the Salesforce managed package, aims to streamline business processes by automating data entry from documents stored in Box, thus reducing manual effort and minimizing errors. This update allows users to upload files to the Box widget, where AI reads and populates relevant Salesforce fields, enhancing data accuracy and workflow speed. The feature includes the Box AI Extract Tab, facilitating intuitive data extraction and context selection for Salesforce objects like Opportunities and Leads. Key applications include lead generation, contract analysis, and invoice processing, with the extended Box for Salesforce Developer Toolkit now supporting Box Hubs for better document organization and collaboration. Additionally, the integration with Salesforce Financial Services Cloud provides compliance teams with audit-ready visibility of document interactions through the new Box Embed and Box Events features. These advancements collectively enhance data integrity and operational efficiency, enabling users to harness AI for faster and more reliable business operations within Salesforce.

Financial institutions face increasing demands for seamless onboarding experiences, yet contend with outdated systems, excessive paperwork, and compliance challenges. The integration of Intelligent Content Management platforms, like Box, offers a solution by digitizing and automating client onboarding processes, thereby reducing turnaround times and enhancing client satisfaction. Box facilitates secure document handling, compliance with regulations such as GDPR and SOX, and integration with over 1,500 applications, including Salesforce and Microsoft 365. AI-enhanced features provide advanced data protection and automate governance, ensuring efficiency and compliance. Real-world applications demonstrate significant productivity gains across banking, insurance, and wealth management sectors, as exemplified by companies like Barnett Capital and USAA, which have achieved measurable improvements in speed and efficiency through Box's platform. By adopting a digital-first approach with Box, financial institutions can modernize operations, streamline workflows, and enhance decision-making, ultimately fostering stronger client relationships and supporting long-term growth.

The text appears to be a detailed excerpt from a JavaScript library or module, likely focused on handling form behaviors and functionalities in web development. It includes functions and methods for manipulating form elements, managing events, handling data, and ensuring compatibility across different browsers and environments. There are also references to interacting with the DOM, managing cookies, and dealing with asynchronous operations. The code snippet seems to be part of a larger framework or plugin, possibly related to form validation, submission, or user interaction handling. Additionally, the text references some specific libraries and technologies like jQuery, indicating that it may be part of a JavaScript-based form handling library or system.

Enterprise AI is encountering a critical challenge as businesses transition from AI experiments to deploying autonomous agents that manage real workflows, discovering that these agents require secure access to business content without creating data silos or security risks. This necessity is being addressed by Box, which has partnered with AWS to offer solutions that facilitate secure, content-aware AI workflows, such as the Box MCP Server and Box MCP + Amazon Quick Suite Solution. These offerings ensure AI agents can access and interact with enterprise content while maintaining governance controls, supported by Amazon Bedrock models to automate complex processes within Box's governed infrastructure. The collaboration between Box and AWS, highlighted at AWS re:Invent 2025 and available on AWS Marketplace, represents a strategic effort to allow AI to operate as a trusted participant in enterprise workflows without compromising established governance foundations.

Box Shield Pro is an AI-powered content protection solution designed to enhance the security of unstructured data within organizations, addressing the increasing challenges posed by advanced threats like ransomware and the explosion of content driven by AI. It introduces features such as automated classification with context-aware AI, which allows for the scalable tagging of sensitive content, and ransomware activity detection, which monitors suspicious file behavior and enables rapid remediation. Additionally, Box Shield Pro includes an AI Threat Analysis Agent that provides concise, actionable threat summaries to streamline security operations and aid in faster triage and response. These enhancements aim to empower security teams by automating content protection processes without hindering user productivity, while also offering advanced capabilities to detect and mitigate potential threats swiftly. Box Shield Pro is part of a broader effort to adapt to the evolving landscape of content security, ensuring organizations can protect their data efficiently as collaboration and content velocity continue to grow.

Box has introduced a comprehensive set of tools for folder manipulation, enabling users to create, rename, move, copy, and delete folders programmatically, enhancing efficiency and scalability in organizing content. These tools support recursive operations for both deletion and listing folder contents, facilitating tasks such as auditing, backup workflows, and content migration. Users can add descriptions and tags to folders, mark them as favorites, and manage access and collaboration settings, providing more control and organization within the platform. These new capabilities allow for improved project management, content migration, and bulk organization by offering features like template setup, access management audits, and automated cleanup. These tools eliminate the need for manual UI navigation, allowing for streamlined workflows and automation, and are now available on the Box Community MCP server.

Anthropic's release of Claude Opus 4.5 marks a significant advancement in AI technology, offering a groundbreaking "effort" parameter that allows users to tailor the AI's reasoning depth to the complexity of a task. Available through Box AI Studio and API, this model enables enterprises to adjust the AI's analytical efforts, providing a balance between precision and speed depending on the stakes of their business decisions. In rigorous evaluations, Opus 4.5 significantly outperformed its predecessor, Opus 4.1, in complex reasoning tasks, achieving higher accuracy in various industry-specific applications such as grant compliance in education, safety regulation analysis in energy, and clinical guideline precision in healthcare. This flexibility in effort levels allows for a high degree of control, making the model particularly useful for high-stakes scenarios where accuracy is critical, as well as for routine tasks requiring quicker responses.