top of page
Search

Two Analyst Reports, One Platform Built for the Area of Agentic AI.

Ryan Rowcliffe
5 minutes ago
8 min read

Gartner IVIP and SACR ARISE


In August, Gartner published an Innovation Insight highlighting the evolution of Identity Visibility and Intelligence Platforms. A short three weeks later, Software Analyst Cyber Research published its inaugural report on Agentic Runtime Identity Security Enforcement.


Two great research reports,  two new categories. 


The cynical read is that analyst firms mint categories the way banks mint products, and some of that is fair. Categories create coverage, coverage creates inquiry volume, inquiry volume creates renewals. I have watched enough named categories arrive and quietly disappear to be skeptical by default.


So be skeptical about this one too, and then notice what makes it different. Neither report describes a new attack or  a new technology. Both describe a gap that has been open for fifteen years, from opposite ends, that is now massively accelerating.


That is not a market inventing a problem. That is two independent groups of analysts realizing  the same market issues and getting ahead of it. 


What Gartner named


Innovation Insight: Identity Visibility and Intelligence Platforms, published inAugust by Rebecca Archambault and Nathan Harris.


Gartner's definition, “a platform that aggregates, normalizes, and analyzes identities, their access, policies, and IAM control configurations across multiple domains, meaning cloud, on-premises, and SaaS. The capabilities named are unified identity posture visibility, continuous risk assessment across disparate systems, access gap identification, and operationalizing mitigation.”

Read that list with fifteen years of IAM history in mind and the omission jumps out. Every capability there is something we have claimed to have since the first identity governance deployment. Aggregate identities, assess risk, find gaps. We bought products for each of these issues in 2012.


The operative words are across multiple domains. Not aggregation within a directory. Aggregation across the cloud estate, the on-prem estate, and the SaaS estate, normalized into one model. That is the part nobody actually built, because each domain arrived with its own tooling and its own console, and integrating them was always next quarter's problem.


Gartner's framing of why it matters now is about AI agents: as enterprises increasingly deploy agentic AI solutions, the need for continuous, intelligent visibility into these identities becomes paramount. Agents are the forcing function, they are not the origin of the gap.


AuthMind was named a Representative Provider, for identity observability across agentic AI, NHIs, and human users spanning cloud, SaaS, and on-premises. Our CEO's line in the release is the one I would keep: “policy intent and actual identity activity are no longer the same thing, and closing that gap requires real observability, not another static control.”


Hold onto that sentence. It is the hinge of this whole post.


What SACR named


ARISE, September 2026, from Paul Webber and Lawrence Pingree with the SACR research team. Roughly thirty vendors surveyed, six market patterns, thirteen profiled in depth.


SACR's focus is the runtime enforcement layer for agents in motion, and the single question at its center is whether this agent should be allowed to take this action, with this tool, credential, data, and delegated authority, for this purpose, right now.


Three layers. Deterministic governance draws the boundary before anything runs. Behavioral and intent analysis watches the sequence. Dynamic runtime governance intervenes in the action itself, with a verb list that runs from block and redact through scope reduction, credential revocation, and termination.


And a six-level scale, Agentic Control Depth, measuring how deep into an agent's action your controls actually reach. ACD 0 is blind execution. ACD 2 is post-event telemetry. ACD 4 is inline intervention before the action completes, which SACR names as the minimum operational threshold for agents touching sensitive data or critical workflows.


Here is where the two reports stop looking like separate conversations.


ARISE layer one is a boundary you cannot draw around a population you have not enumerated. Layer two is drift detection, which requires a baseline, which requires observed history correlated per identity. Both of those are IVIP's job description. SACR did not restate the visibility problem because it was busy defining the enforcement one, but the dependency is structural: you cannot do ARISE completely without IVIP..


And the reverse holds. IVIP tells you the gap between policy intent and actual activity. It does not close the gap. Knowing precisely which agent is doing something it should not, in real time, with excellent fidelity, and having nothing in the path to stop it, is a better-documented version of the same outcome.



One finding, stated twice


Strip the category names off both reports and the same sentence is underneath.


Gartner: what your identity systems are configured to permit and what your identities are actually doing have come apart, across domains, and nobody has a unified view of the difference.


SACR: the gap between configured permission and actual behavior is now exploited faster than a human process can respond, so something has to act inside it.


Now the part I would put in front of anyone treating this as an AI story. That gap did not open in 2026. It opened the first time an organization ran service accounts it could not enumerate, and it has been widening quietly ever since. We knew, every identity architect I have worked with in twenty years could tell you their NHI inventory was incomplete and their service account entitlements were wider than justified.


We accepted it, and the compensating control we accepted it against, whether or not anyone wrote this down, was that exploitation required a human working at human speed, and a human working at human speed would leave enough time for a human on our side to notice.


That was the bet. It was always a worse bet than we admitted, because the second half depended on detection quality we also knew was uneven. Agents did not introduce a new risk. They repriced an old one, and the new price is 22 seconds from initial access to hand-off and one second to cluster admin.


Which explains the timing better than any market theory does. Two research organizations arrived in the same month because the risk acceptance broke in the same time frame, for everybody, at once. Gartner described the measurement gap. SACR described the response gap. They are the two halves of one overdue correction.


The honest version, and the one I would say to a board: this is not a new category of spending. It is a deferred maintenance bill, and the interest came due when the adversary stopped being slow.


Two categories does not mean two products


The procurement risk here is obvious and I have already watched teams walk into it. Two named categories arrive, two line items appear in the FY27 plan, and eighteen months later you own an IVIP product and an ARISE product that share no identity model and join on a service account name.


That is the seam problem, and it is worse in this pairing than most, because the dependency between the two categories is not loose coupling. Enforcement quality is a direct function of what the enforcement point knows about the identity at the moment of decision. Put the knowing in one product and the deciding in another, and you have built a gate that consults a database over the network before it can be selective. This becomes another waste of budget.



Gartner IVIP

SACR ARISE

Published

24 August 2026

September 2026

Question answered

What is actually happening across my identity estate?

Should this action be allowed, right now?

Operates

Continuously, across domains

In the decision path, per action

Output

Posture, risk, access gaps

Allow, block, redact, pause, revoke

Fails as

Accurate and too late

Blind and blunt


Read that bottom row as the buying criteria. The two failure modes are complementary, which means a product strong on one side and weak on the other does not give you half the outcome. It gives you one of the two failure modes with better marketing.


So the question to carry into evaluations is not which category a vendor is in. It is whether one identity model spans both jobs. Does the thing that observes and the thing that enforces share a single record for an agent, including when that agent acts under delegated human authority and presents a service account downstream? If the answer involves the word integration, ask what the join key is and what happens when it drifts.


AuthMind appears in both reports, which is the reason I can write this post, and I will state the conflict of interest plainly rather than bury it. It is also the argument: not that appearing twice is an achievement, but that the two jobs were never separable and building them apart was the mistake.


What I would do Next


Use the vocabulary. That is the practical value of a named category and it is not a cynical point.


Most of you have been asking for money to fix NHI inventory and service account scope for three budget cycles and losing to something with a clearer story. You now have two research organizations, just  weeks apart, describing the gap you have been describing. Rewrite the same proposal in their language and resubmit it. I am entirely serious. The work did not change; the funding argument did.


Audit your identity estate against IVIP's four capabilities and be specific about domains. Can you produce, today, a normalized view of identities and their access across cloud, on-prem, and SaaS in one place? Not four consoles and an analyst who knows how to join them. One view. Most organizations discover the answer is no, and discovering it in a planning meeting is much cheaper than discovering it in an incident.


Score production against ACD, not the pilot, and score it per workload rather than as one number. The gap between what production does and what the pilot demonstrated is usually the whole story.


Then do the thing neither report will do for you: go fix a fundamental. Pick the worst over-scoped service account you already know about, the one everybody has been avoiding because tightening it might break something on a Friday. Tighten it. No new category is required to do that, and no new product will do it for you.


Last, a caution about buying to a category name. Both of these are young. IVIP is an Innovation Insight, which is Gartner's early-stage format, not a Magic Quadrant. ARISE is an inaugural report. Vendor claims of alignment will outpace vendor capability for at least a year, ours included. Test the behavior, not the designation.


A signal, not a strategy


Being named in both reports matters to AuthMind, and I am not going to pretend otherwise. SACR identified us as a representative vendor with broad alignment across all three ARISE layers, one of a small number among the thirteen profiled. Gartner named us a Representative Provider for IVIP. For a company arguing a position the market had not named yet, two independent confirmations in a month is a sign that we have been on the right path..


It is still a signal about a market, not a strategy for your environment. A designation tells you an analyst found a vendor's approach worth describing. It does not tell you whether that vendor can enforce in your decision path, at your latency, across your identity estate. Only a test tells you that.


What I would take from these two reports has nothing to do with vendor lists.


The fundamentals never moved. Least privilege, credential hygiene, segmentation, inventory, continuous verification. All correct fifteen years ago, all correct now, all still mostly unfunded. Neither Gartner nor SACR describes a new defensive principle, because there isn't one.


What both describe, from opposite ends, is the one thing we genuinely never built: correlated observability of identity, human and non-human, across every domain those identities actually operate in. That absence is what let the risk acceptances look reasonable. You cannot price a risk you cannot see, so we priced it at zero and moved on.


Agents did not break anything. They removed the slack that was hiding the bill.


Sources



Gartner does not endorse any vendor, product or service depicted in its re search publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally, and is used herein with permission. All rights reserved. Representative Provider and representative vendor designations do not constitute endorsement.


Comments

bottom of page