Identity at the Center

Jeff and Jim are joined by Rob Ainscough, Chief Identity Security Advisor at Silverfort and host of the Identity Decoded podcast, to look at what happens when an autonomous AI attacker goes after identity. Rob shares how a decade running a 400,000-person IAM program at a UK retailer, and one identity-driven incident, moved him from compliance-led to threat-led identity. He walks through the AI Attack Lab, where an AI red agent called Cypher went from a network foothold to domain dominance in under four minutes by chaining familiar identity weaknesses in parallel. The conversation covers why governance alone can't stop machine-speed attacks, identity as a drawbridge that rises with the threat level, why agents favor non-human identities, and giving every account a protection plan. Plus podcasting lessons, community building, and the return of the Identity Mullet.

This episode is made possible with support from Silverfort. Learn more at silverfort.com/idac

Connect with Rob: https://www.linkedin.com/in/rob-ainscough/

Identity Decoded podcast: https://www.silverfort.com/identity-decoded-podcast/

Connect with us on LinkedIn:

Jim McDonald: https://www.linkedin.com/in/jimmcdonaldpmp/

Jeff Steadman: https://www.linkedin.com/in/jeffsteadman/

Visit the show on the web at idacpodcast.com

00:00 Intro, and an apology to Mark
01:33 Meet Rob Ainscough
01:53 Falling into identity at a UK retailer
03:43 The incident that changed everything
05:48 Orphaned account war stories
07:59 Account correlation in the age of agents
09:23 The Identity Decoded podcast
11:11 What Silverfort has been focused on
13:21 Black October: hype vs. reality
15:51 Familiar techniques, new speed and scale
18:52 What is the AI Attack Lab?
23:23 Watching Cypher compromise a domain
27:58 What identity programs left behind
29:25 Building the attacker and keeping it contained
32:16 Controls that stay transparent to the model
34:08 What surprised Rob most
35:27 Goal-oriented to a fault
37:23 Was the lab environment rigged?
39:16 How security teams react
43:16 Why good governance isn't enough
47:19 Where to start: every account has a plan
49:31 Why AI attackers target NHIs
51:19 Lessons from a first season of podcasting
56:38 Nerves before hitting record
58:27 The Identity Mullet
01:02:38 Wrap-up

IDAC, Identity at the Center, Jeff Steadman, Jim McDonald, Rob Ainscough, Silverfort, Sponsor Spotlight, Identity Decoded, AI Attack Lab, Cypher, AI agents, agentic AI, autonomous attacker, Black October, threat-led identity, identity threat detection and response, ITDR, Active Directory, Kerberoasting, AS-REP roasting, golden ticket, lateral movement, credential harvesting, domain compromise, non-human identities, NHI, service accounts, PAM, IGA, identity debt, runtime enforcement, identity security, CISO, red team, Mythos

Jim McDonald flies solo this week while Jeff Steadman is out on assignment, and he's joined by Autumn Hurley, Cyber Risk Consulting Principal at RSM US LLP, for a practitioner's crash course in risk management frameworks and the three lines model. Autumn shares her path from marketing and banking to running offensive security operations and now leading RSM's cyber financial services team. From there, she and Jim unpack the vocabulary identity teams hear constantly but rarely get explained: likelihood, impact, risk appetite, controls, and the difference between a policy, a standard, and a procedure. They cover what GRC tools do, how frameworks like NIST CSF, ISO 27001, COBIT, and COSO apply, and the warning signs that compliance has become a checkbox exercise. Autumn then walks joiner-mover-leaver through the first, second, and third lines, explains where external audit fits, and offers practical guidance on translating an identity project into the risk language executives, auditors, and boards respond to. The episode closes with home remodeling advice and a contractor story Jim would rather forget.

Connect with Autumn: https://www.linkedin.com/in/autumnhurleyatrsm

Connect with us on LinkedIn:

Jim McDonald: https://www.linkedin.com/in/jimmcdonaldpmp/

Jeff Steadman: https://www.linkedin.com/in/jeffsteadman/

Visit the show on the web at http://idacpodcast.com

00:00 Welcome and upcoming events
02:20 Why this episode: risk and the three lines model
03:36 Autumn's non-traditional path into cybersecurity
05:12 From pen testing to risk and governance
07:18 How her perspective on risk has changed
08:18 What is risk management?
10:07 Risk concepts every identity practitioner should know
12:52 Business, technology, and identity risk as a pyramid
14:28 What is a control?
16:05 Policy vs. standard vs. procedure
17:56 Control design vs. operating effectiveness
20:05 What is a GRC tool?
21:36 Who creates risk and control frameworks
23:07 NIST, ISO 27001, COBIT, COSO: what you need to know
24:24 Keeping compliance from becoming a checkbox exercise
26:16 The three lines model explained
30:19 Where external audit fits
31:38 Joiner-mover-leaver through the three lines
34:04 Can auditors rely on IGA and SSO data?
36:09 Why the lines are blurring
37:50 The IT risk function as internal advisor
38:46 Preparing for risk conversations with senior leaders
40:40 Translating an identity project into risk language
42:36 Key terms for talking with risk, compliance, and audit
43:56 Autumn's key takeaway
45:38 Remodeling advice and a contractor horror story
49:37 Wrap-up

IDAC, Identity at the Center, Jeff Steadman, Jim McDonald, Autumn Hurley, RSM, risk management, three lines model, three lines of defense, cyber risk, risk management framework, controls, control effectiveness, policy, standards, procedures, GRC, governance risk and compliance, NIST CSF, ISO 27001, COBIT, COSO, internal audit, external audit, risk appetite, risk tolerance, likelihood and impact, joiner mover leaver, JML, user access reviews, IGA, identity governance, privileged access, financial services, compliance, IAM, identity and access management

Jeff and Jim finally land a long-sought pair of guests: current Gartner analysts Rebecca Archambault and Paul Rabinovich, chair and co-chair of the Gartner Identity & Access Management Summit, December 7-9 at Caesars Forum in Las Vegas. Rebecca and Paul share their paths into identity, from HSBC and Lotus Notes to a security PhD and aerospace and defense, then preview this year's theme of reshaping identity in the age of AI. The conversation covers the MCP interop session with the OpenID Foundation, why authorization remains so hard to standardize, runtime controls, human-in-the-loop, and intent-based access control for AI agents, IVIP and the end of "light IGA," and why the fundamentals still matter. They close with tips for first-time attendees, how analyst one-on-ones work, the new expo floor neighborhoods, and a few Las Vegas recommendations.

Gartner US IAM Summit: https://www.gartner.com/en/conferences/na/identity-access-management-us

Connect with Rebecca: https://www.linkedin.com/in/rebecca-becky-archambault-4b4285111/

Connect with Paul: https://www.linkedin.com/in/paul-rabinovich-20b79b8/

Connect with us on LinkedIn:

Jim McDonald: https://www.linkedin.com/in/jimmcdonaldpmp/

Jeff Steadman: https://www.linkedin.com/in/jeffsteadman/

Visit the show on the web at http://idacpodcast.com

00:00 Jeff's broken finger update and the Gartner white whale
03:17 Welcome Rebecca Archambault and Paul Rabinovich
04:08 Rebecca's path from HSBC to identity
06:04 Lotus Notes war stories
07:23 Paul's path from a security PhD to identity
08:22 Paul's role in Gartner for Technical Professionals
09:27 Rebecca's coverage: IAM programs, IGA, IVIP, ISPM, ITDR, and CIAM
10:43 How their perspectives changed as analysts
12:50 What to expect at the Gartner IAM Summit
14:54 What's different this year
19:17 The MCP interop with the OpenID Foundation
23:32 Past interops: CAEP and AuthZEN
24:09 What didn't make the agenda
26:36 Decentralized identity and the London Summit
28:32 Rebecca's session: Renaissance in the Age of AI
31:58 Can authorization ever be standardized?
39:42 Standards and authorizing AI agents
42:12 Paul's session: AI agents on a solid access management foundation
46:50 Foundational IAM and the rise of IVIP
49:23 First-timer tips: Conference Navigator, recordings, one-on-ones
55:21 Las Vegas hidden gems
1:00:31 PSA: Caesars Forum is not Caesars Palace
1:04:46 New expo floor neighborhoods
1:05:56 Networking advice
1:07:15 Wrap up: Ask IDAC, IAM 101 series, and the donor page

IDAC, Identity at the Center, Jeff Steadman, Jim McDonald, Rebecca Archambault, Paul Rabinovich, Gartner, Gartner IAM Summit, Identity and Access Management, IAM, AI Agents, Agentic AI, MCP, Model Context Protocol, OpenID Foundation, Interop, Authorization, AuthZEN, CAEP, Shared Signals Framework, OPA, Cedar, Intent-Based Access Control, Human in the Loop, IGA, IVIP, ISPM, ITDR, CIAM, PAM, Machine Identity, NHI, Zero Trust, Post-Quantum Cryptography, Decentralized Identity, Las Vegas

In this Sponsor Spotlight, Jim McDonald talks with Pranay Yadav, co-founder and CEO of Iden, about why identity governance still runs on tickets and spreadsheets at so many organizations. Pranay came to identity from venture capital at Accel, where he kept hearing the same story from IT teams. SSO arrives early, and everything after it gets tracked by hand.

They dig into how deep governance has to go. Knowing someone has a GitHub account is one thing. Knowing they can approve their own pull requests is another. Pranay explains where SCIM stops short of fine-grained entitlements in SaaS, internal, and legacy apps, why a governance program is only as complete as the weakest app in the stack, how teams can sort apps into P0, P1, and P2 priorities, and when access requests should be automated versus kept in front of a human.

The conversation closes on AI. That covers using it to help practitioners and end users, the harder problem of governing AI agents that share a person's identity, credentials, and long-lived tokens, and why nobody has patience for multi-year identity programs anymore. Pranay also shares the advice he'd give his firstborn: don't typecast yourself.

Made possible with support from Iden. Learn more at idenhq.com/idac

Connect with Pranay: https://www.linkedin.com/in/ypranay/

Connect with us on LinkedIn:

Jim McDonald: https://www.linkedin.com/in/jimmcdonaldpmp/

Jeff Steadman: https://www.linkedin.com/in/jeffsteadman/

Visit the show on the web at idacpodcast.com

00:10 Welcome and Sponsor Spotlight introduction
01:09 From venture capital to identity governance
02:41 Where the friction starts for small IT teams
04:37 Starting with the apps that lack built-in provisioning
05:21 The floor and the ceiling of who feels the pain
06:44 How deep governance has to go
11:35 When knowing "has access" is enough
13:37 Where SCIM fits and where it stops
15:18 Depth in practice: SaaS roles, GitHub, and internal apps
19:13 Breadth and the long tail of apps
21:30 Prioritizing apps into P0, P1, and P2
25:39 What governance means functionally
28:31 When to automate access requests and when to keep a human in the loop
32:18 The experience for the IGA administrator
36:25 AI for governance
43:06 Why governing AI agents is so hard
47:41 No appetite for multi-year identity projects
52:49 Lighter note: advice for a firstborn
56:05 Wrap-up

IDAC, Identity at the Center, Jeff Steadman, Jim McDonald, Pranay Yadav, Iden, Sponsor Spotlight, identity governance, IGA, identity governance and administration, IAM, SCIM, fine-grained entitlements, RBAC, authorization, access reviews, joiner mover leaver, onboarding, offboarding, provisioning, deprovisioning, least privilege, segregation of duties, audit evidence, connectors, long tail of applications, SaaS governance, internal applications, legacy applications, automation, human in the loop, AI for governance, MCP, AI agents, agent governance, non-human identity, delegated identity, API tokens, Accel

Recorded live from the expo hall at InfoSec World 2026, Jeff and Jim welcome back Joseph Carson, host of the Security by Default podcast, for his third appearance on the show. Joseph explains how his thinking has moved from "AI vs AI" to "AI vs reality" after talking with CISOs at roughly 40 events this year, including stories of agents breaking into systems they were never meant to touch and a CFO vibe coding a family scheduler on a corporate laptop. He walks through his three-tier framework for agent risk (task, operational, and legally binding agents), makes the case that agents need identifiers and audit trails rather than human-style treatment, and describes Identity Under Attack, a Dungeons and Dragons style tabletop exercise he ran at NATO CyCon with identity coins, chaos cards, and dice. The conversation also covers why identity resiliency keeps landing last on the shopping list, what Locked Shields teaches about communication and identity strategy, a sobering prediction for 2027, and what identity practitioners can do today with ephemeral, auditable access. The episode closes with a podcaster gear chat and the difference between AI slop and AI enabled work.

Connect with Joe: https://www.linkedin.com/in/josephcarson/

Security by Default podcast: https://www.youtube.com/@SecuritybyDefaultPodcast

Connect with us on LinkedIn:

Jim McDonald: https://www.linkedin.com/in/jimmcdonaldpmp/

Jeff Steadman: https://www.linkedin.com/in/jeffsteadman/

Visit the show on the web at http://idacpodcast.com00:00 Intro from InfoSec World 2026 and Jeff's broken finger
02:33 Welcome back Joseph Carson
03:18 From AI vs AI to AI vs reality
05:02 CISO stories: rogue agents and the vibe-coding CFO
06:46 The rise of consumer agents
07:50 A three-tier framework for agent risk
10:52 Onboarding digital workers with a code of conduct
12:39 Identities or identifiers? Estonia's agent identity debate
15:34 Identity Under Attack: a Dungeons and Dragons style tabletop exercise
22:32 Bringing the exercise to conferences
24:49 What surprised Joseph: collaboration across roles
26:52 Common choices and why MFA comes first
28:16 Designing a budget that forces hard choices
29:29 Identity resiliency as an afterthought
31:24 AI as the Mario Kart mushroom
32:53 Lessons from NATO Locked Shields
35:49 Supply chain and third-party risk
37:18 Predictions for 2027
38:59 What identity practitioners can do now
40:23 Vibe coding and governance at AI speed
42:04 Podcast gear and AI-assisted production
48:02 AI slop vs AI enabled
52:17 About the Security by Default podcast
53:45 Wrap-up

IDAC, Identity at the Center, Jeff Steadman, Jim McDonald, Joseph Carson, Security by Default, InfoSec World 2026, Cyber Risk Alliance, agentic AI, AI agents, non-human identity, agent identity, digital workers, AI governance, vibe coding, tabletop exercise, Identity Under Attack, gamification, NATO CyCon, Locked Shields, MITRE ATT&CK, incident response, identity resiliency, MFA, privileged access, just-in-time access, ephemeral access, least privilege, supply chain risk, third-party risk, AI slop, podcasting