Microsoft fixed a dangerous role misconfiguration in Entra that briefly allowed broad takeover of service principals across entire tenants. This episode explains what happened and why firms should treat service principal ownership with the same seriousness as high‑privilege user accounts.

For founders and firm leaders, the deeper issue is the pace at which cloud platforms introduce new roles and preview features. These changes can expand exposure quietly, and traditional access reviews often fail to keep up. We break down what to check now and how to tighten your monitoring.

We also cover the Cohere and Aleph Alpha merger, a long‑standing Linux privilege escalation bug, and Infor's view on why data security is slowing enterprise AI programs.

Learn more at https://crestvale.io

Support the show

Today's episode looks at how agentic AI is shifting from experimentation to billable work. Google, Microsoft, and leading security vendors are signaling that firms should turn pilots into repeatable services, with early movers setting the price points clients will adopt.

This matters because clients are no longer paying for AI concepts. They are paying for deployment, governance, assurance, and proof that systems are doing what they should. Firms that act now will shape expectations and keep the client relationship anchored to measurable results.

We also cover why CFOs are treating AI and workflow automation as the strongest remaining cost lever, how leading firms are filtering vendor hype with disciplined pilots, and why recent Teams‑based attacks show that internal chat has become an active risk surface.

Learn more at https://crestvale.io

Support the show

Google is launching the Gemini Enterprise Agent Platform, a unified control system for building and managing AI agents across an organization. The new platform replaces scattered tools with a single place to handle development, data access, scaling, and governance. For firms planning to operationalize agents, this move sets new expectations for what enterprise AI needs to include.

This matters because firms are shifting from pilot projects to real workflows that must stay reliable and compliant. Long-running agents, persistent memory, and stronger oversight will shape how professional service firms automate their work and how quickly they can deploy internal and client-facing tools.

The episode also covers the Freshfields and Anthropic partnership, new warnings about China-linked router botnets, and Singapore's operational playbook for governing agentic AI.

Learn more at https://crestvale.io

Support the show

This episode looks at the growing demand for operational AI governance. Acronis introduced a new service that turns AI oversight into a recurring managed offering, giving firms real visibility into client use and the ability to enforce guardrails before problems develop.

For professional service leaders, the shift matters because clients are adopting AI faster than internal controls can keep up. This creates risk, but it also opens a new service category for firms that want to lead on responsible AI use instead of reacting to exposure after the fact.

We also cover the collapse of the accounting talent pipeline, new signs that BigLaw attrition is tied to weak AI enablement, and the latest move in Congress to narrow Corporate Transparency Act reporting.

Learn more at https://crestvale.io

Support the show

Today's episode breaks down the surge in AI driven security incidents across professional service firms. AI agents are now showing up inside environments without approval, oversight, or controls, and the operational impact is already visible. Firms are facing data exposures, disruptions, and unintended actions because bots with real credentials are making decisions no one is monitoring.

This matters for firm leaders because AI agents now behave like users. They hold access, trigger workflows, and shape outcomes. Without clear onboarding, auditing, and offboarding, they create blind spots that attackers and regulators can exploit. Firms that build discipline around agent governance will reduce incidents and avoid regulatory trouble.

We also cover new findings from Kroll, board level responses to fast moving frontier models, and the rising GDPR risks tied to AI in internal investigations.

Learn more at https://crestvale.io

Support the show

Today's episode breaks down the breach at Vercel that started with a single employee connecting a consumer AI app to a corporate account. It is a clear example of how everyday tools create hidden access paths that can bypass even strong security programs.

For firm leaders, the message is direct. OAuth permissions are now a frontline risk. If teams are linking personal AI tools or browser extensions to Google Workspace or Microsoft accounts, the exposure is already in play. This episode explains what happened and how to assess your own environment.

We also cover Adobe's new agent driven customer experience layer, collapsing AI adoption rates inside firms, and how BBVA turned shadow AI into a managed enterprise system.

Supporting stories include Booking dot com's security reset, new DNS guidance from NIST, and growing malware threats targeting operational systems.

Learn more at https://crestvale.io

Support the show

AI driven cyberattacks are accelerating, and boards are being warned that security can no longer sit in an IT corner. This episode breaks down new guidance from the United Kingdom, fresh model releases from OpenAI, and the latest breach hitting a major hosting provider. The theme running through all of it is simple: attack speed is rising, and governance needs to keep up.

Professional service firms are sitting in a risk zone. Clients are adopting AI assisted defense. Regulators are raising expectations. And attackers are exploiting weaknesses in supply chains and hosting platforms. Leaders who treat this as a strategic risk will stay aligned with client expectations and protect their operating base.

We also cover supporting developments across AI vendor curation, hybrid cloud strategy, and network level fraud detection.

Learn more at https://crestvale.io

Support the show

This episode breaks down the surge in MFA bypass phishing kits after the Tycoon takedown. The removal of more than three hundred domains did not slow attackers. It pushed them into smaller platforms that copied Tycoon's methods and expanded them. For firm leaders, this shift signals that MFA alone no longer blocks modern credential theft.

This matters because these new kits intercept one time codes and session tokens in real time. Firms that rely on older authentication assumptions are exposed, and many have not updated governance or controls to match the current threat landscape. This episode explains the practical changes leaders need to understand.

We also cover the new FortiSandbox exploit, the growing need for AI governance in cyber programs, and how solo operators are replacing junior hires with AI agents.

Learn more at https://crestvale.io

Support the show

New York regulators have tightened cybersecurity attestations, creating immediate downstream pressure on every service firm that works with clients under state oversight. This episode explains what changed, why the new multi factor authentication and asset inventory requirements matter, and how they will influence vendor expectations across accounting, law, and advisory practices.

For firm leaders, the impact is significant. Compliance risk now extends into your own internal systems and controls, whether you are ready or not. The episode breaks down the practical implications and the decisions firms will face as clients align with the new standard.

We also cover Zoom's new human verification checks, Zenskar's funding for complex revenue automation, and the ongoing disconnect between CEOs and CISOs.

Learn more at https://crestvale.io

Support the show

American Express is buying Hyper to pull AI agents directly into corporate spend workflows. This move signals a shift in how expense management will work, pushing policy enforcement and review into the moment a purchase happens instead of the end of the month.

For professional service firms, this matters because manual expense workflows will soon feel slow and error‑prone compared with automated systems. The firms that modernize early will gain tighter controls and cleaner data with less back‑office labor. Those that wait will fall behind as clients and staff expect faster cycles.

This episode also covers new automation in immigration practices, rising cyber losses linked to AI and crypto, and how offshoring continues to reshape firm economics.

Learn more at https://crestvale.io

Support the show