AI Advances
AI agents and tools are changing the game for security and infrastructure
Building Better AI Agents
Brex has been making waves in the AI world with its innovative approach to building AI agent policy. According to VentureBeat, Brex built its AI agent policy by watching what agents actually do, not by writing rules first. This approach has led to the development of an internal platform called CrabTrap, an open-source HTTP/HTTPS proxy that intercepts all network traffic, examines policy rules, and uses a large language model to make decisions.
This approach is significant because it addresses the limitations of traditional guardrails in containing what agents are doing with real credentials like API keys, OAuth tokens, and service accounts. As OpenClaw has become one of the most widely adopted agentic frameworks, Brex's approach could be a game-changer for enterprise-scale AI adoption.
For context, AI agents are software programs that can perform tasks autonomously, and they require real credentials to work effectively. However, these credentials can be a security risk if not managed properly. Brex's approach to building AI agent policy is a step towards mitigating this risk and making AI agents more secure and reliable.
The implications of Brex's approach are significant, and it will be interesting to see how other companies adopt similar strategies. As AI agents become more prevalent, the need for secure and reliable management of these agents will become increasingly important. Brex's approach could be a model for other companies to follow, and it will be exciting to see how this technology evolves in the future.
Capital One's VulnHunter
Capital One has released VulnHunter, an open-source, agentic AI security tool that scans source code for exploitable vulnerabilities, maps out how an attacker would reach them, and proposes targeted fixes. According to VentureBeat, this tool is one of the most ambitious attempts by a major financial institution to turn offensive AI capabilities into a public defensive resource.
The release of VulnHunter is significant because it addresses the growing need for security teams to identify and fix vulnerabilities before hackers can exploit them. As AI-powered attacks become more common, the need for AI-powered defense is becoming increasingly important. Capital One's release of VulnHunter is a step towards making AI-powered defense more accessible to companies of all sizes.
For context, security teams are facing a rising tide of threats, and the traditional approach to security is no longer sufficient. AI-powered security tools like VulnHunter are becoming increasingly important for identifying and fixing vulnerabilities before they can be exploited. Capital One's release of VulnHunter is a significant contribution to this effort.
The implications of VulnHunter are significant, and it will be interesting to see how other companies adopt similar strategies. As AI-powered attacks become more common, the need for AI-powered defense will become increasingly important. Capital One's release of VulnHunter could be a model for other companies to follow, and it will be exciting to see how this technology evolves in the future.
Closing the Gap between AI Agents and Legacy Infrastructure
Legacy infrastructure is slowing down AI agents, not the models themselves. This was the shared conclusion of three infrastructure leaders from LinkedIn, Walmart, and Zendesk at VB Transform 2026. According to VentureBeat, the panel brought together Animesh Singh, senior director of AI platform and infrastructure at LinkedIn, Desiree Gosby, SVP of corporate technology services and technology strategy at Walmart, and Sami Ghoche, VP of applied AI at Zendesk.
Each of the panelists described what actually broke when they moved agents from pilot to production, and they all arrived at the same conclusion: legacy infrastructure is the bottleneck. This is significant because it highlights the need for companies to invest in modernizing their infrastructure to support AI agents.
For context, AI agents require fast and reliable infrastructure to operate effectively. However, many companies are still using legacy infrastructure that is not designed to support the needs of AI agents. This can lead to slowed performance, increased latency, and decreased reliability.
The implications of this conclusion are significant, and it will be interesting to see how companies respond. As AI agents become more prevalent, the need for modern infrastructure will become increasingly important. Companies that invest in modernizing their infrastructure will be better positioned to take advantage of the benefits of AI agents.
Databricks' Valuation and AI Research
Databricks has remade its image into an AI company and has published research on the cost savings of open weight AI models for coding. According to TechCrunch, Databricks has hit a valuation of $188B, extending its run as AI's favorite second act.
This is significant because it highlights the growing importance of AI in the tech industry. Databricks' research on the cost savings of open weight AI models for coding is also significant because it highlights the potential for AI to reduce costs and improve efficiency in software development.
For context, AI models are becoming increasingly important in software development, and the cost of training and deploying these models can be significant. Databricks' research on open weight AI models for coding highlights the potential for AI to reduce costs and improve efficiency in software development.
The implications of Databricks' valuation and research are significant, and it will be interesting to see how the company continues to evolve and innovate in the AI space. As AI continues to grow in importance, companies like Databricks will be at the forefront of this trend.
Agility Robotics' New Training Center
Agility Robotics is opening a new training center for its Digit robots in Fremont, California. According to TechCrunch, this move is significant because it highlights the growing importance of robotics and AI in the tech industry.
The new training center will allow Agility Robotics to train and deploy its Digit robots more effectively, and it will also provide a hub for innovation and research in the field of robotics and AI.
For context, robotics and AI are becoming increasingly important in a variety of industries, from manufacturing to healthcare. Agility Robotics' new training center highlights the potential for robotics and AI to improve efficiency and productivity in these industries.
The implications of Agility Robotics' new training center are significant, and it will be interesting to see how the company continues to innovate and evolve in the field of robotics and AI. As robotics and AI continue to grow in importance, companies like Agility Robotics will be at the forefront of this trend.
The bottom line
The stories highlighted above demonstrate the growing importance of AI and its applications in various industries. From building better AI agents to closing the gap between AI agents and legacy infrastructure, these stories highlight the need for companies to invest in AI research and development.
The implications of these stories are significant, and it will be interesting to see how companies respond to the growing demand for AI-powered solutions. As AI continues to evolve and improve, it will be exciting to see how it changes the world.
- AI agents and tools are changing the game for security and infrastructure
- Legacy infrastructure is slowing down AI agents, not the models themselves
- Databricks' valuation and research highlight the growing importance of AI in the tech industry
- Agility Robotics' new training center highlights the potential for robotics and AI to improve efficiency and productivity in various industries
- Companies must invest in AI research and development to stay ahead of the curve
π Built by Mapt
Like this site? Mapt builds websites, brands & growth engines β over text.
π Full episode transcript
Brex just built an AI agent policy by watching what agents actually do, not by writing rules first, and it's a game-changer for enterprise security. This approach is significant because traditional guardrails couldn't contain what those agents were doing with real credentials like API keys, OAuth tokens, and service accounts. Brex set out to overcome these limitations by building an internal platform it calls CrabTrap, which learns from the agents themselves. By doing so, Brex has been able to create a more effective and adaptive security system, one that can keep up with the ever-changing landscape of AI agents. This move is crucial for companies that are increasingly relying on AI agents to perform critical tasks, as it helps to mitigate the risks associated with granting these agents access to sensitive data and systems.
The implications of Brex's approach are far-reaching, and it could potentially change the way companies think about AI security. By focusing on what agents actually do, rather than trying to write rules for them, Brex is able to create a more nuanced and effective security system. This is especially important as AI agents become more widespread and are given more autonomy to make decisions and take actions. As we move forward, it will be interesting to see how other companies adopt similar approaches to AI security. Moving on, another major development in the AI security space is the release of VulnHunter, an open-source AI tool by Capital One that finds software flaws before hackers do.
VulnHunter is a significant development because it uses agentic AI to scan source code for exploitable vulnerabilities, map out how an attacker would reach them, and propose targeted fixes. This tool has the potential to revolutionize the way companies approach software security, by identifying and addressing vulnerabilities before they can be exploited. The fact that Capital One is releasing VulnHunter as open-source is also noteworthy, as it will allow other companies to build upon and improve the tool. By making VulnHunter available to the public, Capital One is helping to create a more secure software ecosystem, and setting a high standard for other companies to follow. Speaking of AI and infrastructure, a recent panel at VB Transform 2026 highlighted the challenges of integrating AI agents with legacy infrastructure.
The panel, which included leaders from LinkedIn, Walmart, and Zendesk, discussed how legacy infrastructure is often the bottleneck when it comes to deploying AI agents. This is because agents think in milliseconds, but legacy infrastructure can't keep up. The panelists shared their experiences of what broke when they moved agents into production, and how they had to adapt their infrastructure to support the speed and agility of AI agents. This is an important reminder that the successful deployment of AI agents requires not just advances in AI technology, but also significant investments in infrastructure. As companies continue to adopt AI agents, they will need to prioritize infrastructure upgrades to ensure that they can support the speed and agility of these agents. In other news, Databricks has hit a valuation of $188 billion, extending its run as AI's favorite second act.
Databricks' success is a testament to the growing importance of AI in the tech industry, and the company's ability to adapt and evolve in response to changing trends. By remaking its image into an AI company and publishing research on the cost savings of open weight AI models for coding, Databricks has positioned itself at the forefront of the AI revolution. The company's focus on AI has clearly paid off, and its valuation is a reflection of its success. Finally, Agility Robotics is planting its flag in Tesla's backyard, opening a new training center for its Digit robots in Fremont, California. And that's all for today, tune in tomorrow when we'll be discussing how Google's new AI-powered chip is set to disrupt the entire semiconductor industry.