AI Research

Box CIO, Heather Ceylan, shares her 10 key takeaways from this year's RSA Conference.

COO Olivia Nottebohm shares real-world lessons from the Box AI transformation.

AI pilots in state and local government fail 67% of the time—but it doesn't have to be that way. Learn the five strategies government leaders are using to move AI from experimentation to real results while maintaining security and public trust.

AI-powered attackers now operate continuously, probing systems at machine speed and chaining vulnerabilities into critical attack paths—while most organizations still rely on periodic penetration testing. We're shifting to continuous adversarial…

BTS is showing what it really takes to turn AI from a promising pilot into lasting operational change. By embedding AI directly into everyday workflows with Box as its secure content foundation, the company is improving efficiency, governance, and real…

As software development accelerates to machine speed, traditional security architecture reviews have become a manual bottleneck that no longer scales. This post explores how we are transforming threat modeling into machine-consumable security…

Ravi Malick's February 2026 newsletter: Why Moltbook's AI chaos is the blueprint for what enterprise agent security must never become.

AI is redefining vulnerability management by shifting the focus from managing an endless backlog of tickets to achieving true risk reduction. In part 3 of our series, Heather Ceylan shares how we are using context-aware prioritization and automated…

Owens Corning successfully converted decades of handwritten lab notebooks and test results into a searchable AI-powered knowledge base using Box. Now, employees can instantly query 85 years’ worth of research data, eliminating time-consuming manual…

In Part 2 of this five-part series, Box CISO Heather Ceylan shares how traditional security detection engineering struggles to keep pace with cybersecurity.

Model evaluation of OpenAI's new GPT 5.4

In this episode of our Partner Podcast Series, Box CTO Ben Kus and Braintrust unpack why AI that grades itself beats AI that just generates.

In Part 1 of this five-part series, Box CISO Heather Ceylan shares how AI is transforming the SOC—shifting analysts from reactive alert triage to proactive threat hunting.

Box anchors persistent agentic memory in a secure, governed cognitive vault, transforming vulnerable local files into auditable, resilient AI.

Box CIO Ravi Malick’s January newsletter highlights why unified data governance is the essential leadership priority for navigating the next wave of European AI regulation.

Box tested Google’s new Gemini 3.1 Pro model in Box AI Studio and found a 6-point accuracy boost on complex enterprise tasks.

Box's legal team slashed contract review time from 8 hours to 1 hour using AI-powered contract management—and now they're sharing the playbook.

Box CCO Jon Herstein talks with RWS Global COO Jake McCoy on the AI-First Podcast about how the entertainment giant uses AI to free humans to do what they do best: create.

Box tested Anthropic’s Claude Sonnet 4.6 on complex, real-world enterprise workflows, and it’s now available in Box AI Studio to power more reliable, high-precision agents and automation.

Claude Opus 4.6 in Box AI delivers a major reasoning upgrade that transforms complex enterprise tasks from tedious manual work into automated, expert-level workflows.