OpenAI adds Lockdown Mode for ChatGPT

OpenAI adds Lockdown Mode for ChatGPT

AI tools are forcing a new tradeoff between capability and control. OpenAI's Lockdown Mode makes that explicit by limiting what ChatGPT can access during sensitive work, rather than trying to eliminate risk entirely. For professional service firms, this shifts AI from a productivity tool into a governance decision. Leaders now need clear policies for when full capability is acceptable and when restricted environments are required. At the same time, runtime AI governance, stricter cybersecurity laws, and the move away from VPNs toward Zero Trust are raising the bar on what "secure" actually means in client work. We also cover new signals from Wallarm, Canada's Bill C-8, and a growing push toward Zero Trust access in law firms. Learn more at https://crestvale.io Support the show [https://www.buzzsprout.com/2602483/support]

Fake IT staff hit law firms in-person

Physical access is becoming the new attack vector for professional service firms. Today's episode looks at the rise of ransomware groups showing up in person at law offices, bypassing traditional cybersecurity defenses entirely. For firm leaders, this shifts the problem from technical controls to operational discipline. Identity verification, front desk protocols, and staff awareness now sit at the center of risk management. At the same time, CMMC enforcement is tightening through contract pressure, and firms that cannot produce evidence of compliance are already being filtered out of revenue opportunities. Meanwhile, platforms like Filevine are racing to become the operational brain of the firm through AI-driven workflows. We also cover changes from the IRS, signals from AI infrastructure markets, and rising pressure on security and software spend. Learn more at https://crestvale.io Support the show [https://www.buzzsprout.com/2602483/support]

Trump AI EO makes patching a compliance issue

AI security just became an operational requirement, not a policy discussion. New federal direction is pushing vulnerability management and rapid patching into enforceable territory, with implications that extend well beyond large tech companies. For professional service firms, this shift will show up in client demands, audits, and engagement terms. The ability to prove disciplined security practices is quickly becoming a prerequisite for winning and keeping work. At the same time, leading firms are productizing their expertise, compressing delivery timelines, and changing how services are priced and delivered. We also cover Kirkland and Ellis partnering with Palantir, a fast-moving developer security exploit, and Aprio's continued push toward a multidisciplinary firm model. Learn more at https://crestvale.io Support the show [https://www.buzzsprout.com/2602483/support]

Ramp Stack launches agentic close for accounting

Automation is moving from assistance to execution inside accounting firms. Ramp's new Stack platform signals a shift where AI agents can run the monthly close end to end, with auditability built in. That changes how work gets done and how firms price it. For firm leaders, this is not just another tool. It challenges the labor model behind core revenue. At the same time, risks are expanding at the identity layer, large firms are acquiring implementation capability, and platforms are opening to direct agent access. The common thread is clear: control is shifting toward systems that execute, not just advise. Also covered: a Microsoft 365 mobile security concern, Grant Thornton's acquisition strategy, Morgan Stanley opening platforms to AI agents, and key signals from GitLab, Coralogix, Google, Meta, and Uber. Learn more at https://crestvale.io Support the show [https://www.buzzsprout.com/2602483/support]

Workday launches Agent Passport for AI verification

AI is moving faster than the systems designed to control it. Today's episode focuses on how governance, verification, and security are becoming the real constraints as firms adopt AI inside sensitive environments. Workday's new Agent Passport signals a shift from building AI to proving it is safe. At the same time, Cisco and Anthropic are accelerating the pace of vulnerability discovery and response, forcing firms to rethink how they handle patching, monitoring, and vendor risk. The result is a new operating reality where speed without control creates exposure. We also cover a major supply chain attack tied to Red Hat packages and what it reveals about CI pipeline risk, along with key moves in M&A, SaaS financing, and AI-related litigation. Learn more at https://crestvale.io Support the show [https://www.buzzsprout.com/2602483/support]