TL;DR
Security leaders are held accountable for an outcome nobody has given them a way to measure. The prevailing model is retrospective and brutal: there was a breach, so you're gone.
AttackIQ has repositioned from breach and attack simulation into full continuous threat exposure management, spanning all five phases, and acquired DeepSurface for attack path discovery and risk scoring.
The interesting artifact is a quantifiable threat debt index, backed by an impact-weighted probability of breach model. It's an attempt to give security a number that behaves like a business metric.
The harder problem sits outside any product. A platform can identify exposures perfectly and still produce zero risk reduction, because identifying an exposure and having the authority to make somebody fix it are different things.
Buying a capability is not the same as running a program, and the industry has been quietly conflating the two for years.
Every security tool eventually collides with the same wall: it can tell you what's wrong, and it cannot make your organization act.
AttackIQ's bet is that the way through is measurement, giving leadership a number that tracks program effectiveness rather than a list of findings nobody owns. Whether the number holds up is the open question. That it's the right question seems hard to dispute.
We sat down for a conversation last week with AttackIQ CEO Brett Galloway at Black Hat USA, and it was by far one of the most memorable and meaningful discussions we had.
A Word on Why This Conversation Went the Way It Did
There's a particular kind of conversation you only get with a leader who came up as a technologist and a practitioner. The layers come off early. You skip the positioning, land in the architecture within a couple of minutes, and somewhere shortly after that the language gets a little saltier than any press release would allow.
Those are our favorite conversations, and they're the ones GigaOm is built for, because our analysts came out of the same chairs. We've stood up the programs, run the tools, and been the person in the room who couldn't get anyone to fix the thing. When the person across the table has done that too, you stop performing and start comparing notes.
Brett's passion was evident from the first minute, and as a team we loved every minute of it.
Let me also just say it plainly: Brett Galloway is a badass. Thirty-plus years in, an engineering degree from Stanford he still visibly uses, a company he took through an IPO behind him, and he would rather argue architecture with you than recite a positioning statement.
People who worked for him years ago still talk about him with real warmth, which in this industry tells you considerably more than a bio does. You do not often get a CEO who will go straight at the uncomfortable parts of his own market, including the parts where his product isn't the answer.
Some Background About AttackIQ
Quick orientation, because their category has moved.
AttackIQ built its reputation on breach and attack simulation: technical validation of whether your security controls actually do what you believe they do, with tight alignment to the MITRE ATT&CK framework.
That's still the engine. But they've repositioned as a full continuous threat exposure management (CTEM) provider covering all five phases, from scoping and discovery through prioritization, validation, and visualization.
The acquisition that made that possible was DeepSurface, which brought attack path graphing and a risk scoring engine. That matters more than a typical tuck-in, because it changes the question the platform can answer. Validation tells you whether a control works. Attack path analysis tells you which broken thing an attacker can actually chain into something that matters. Those are different questions, and most vendors only answer one.
The company employs around 130 people, is cash flow positive, and is growing, which in a market full of companies burning capital to buy category presence is worth noting.
The Measurement Problem
Here's the observation that made the conversation worth having, and it isn't really about AttackIQ's product at all.
Brett's position is that the core problem for security leaders is that they cannot measure the business outcome they're responsible for. The outcome is not getting breached. There is no widely accepted instrument for that. So accountability defaults to something retrospective and crude: something bad happened, so somebody loses their job.
Sit with how strange that is compared with any other executive role:
A CFO has closing figures
A revenue leader has pipeline and attainment
An operations leader has uptime and throughput
All imperfect, all gameable, all considerably better than nothing.
The security leader gets a stack of findings, a spend number, and an implicit promise that if the year passes quietly, they were probably doing it right.
AttackIQ's answer is a threat debt index: a quantifiable metric intended to measure, track, and optimize the effectiveness of a security program over time, sitting on top of an impact-weighted probability of breach model.
I like the framing considerably more than I expected to. Naming it debt is doing real work, because debt is a concept every executive already understands. It accumulates, it carries interest, you can service it deliberately or let it compound, and you can compare this quarter with last. The analogy to technical debt is obvious, and that's precisely why it travels. Nobody needs the concept explained.
Whether the underlying model is sound is a separate question, and one worth pressing on. But the ambition is correct. If security is going to be governed like a business function, it needs a number that behaves like a business metric.
There's something else worth drawing out here, and it has less to do with the metric than with who was saying it.
This came from the most senior person in the company. Not a product marketer working from a messaging document, not a field engineer describing a feature, but the CEO, articulating unprompted that the fundamental problem in his market is that his buyers are accountable for something they cannot measure.
That should make any practitioner CISO and their team meaningfully more comfortable about how AttackIQ approaches this space.
When the person setting direction understands the actual job, the roadmap tends to follow. When they don't, you get products built to win feature comparisons and demos.
A CEO who can describe your problem better than your own executive summary does is a CEO whose engineering priorities are probably pointed somewhere useful.
For buyers, it also means the hard conversations are available to you.
If you take an uncomfortable question to this company, you're not going to get handled. That's worth a great deal at renewal time, and it's not something you can tell from a datasheet.
The Bottleneck No Product Can Fix
This was the deepest part of the conversation, and the most honest.
A platform can map your attack paths perfectly, score your exposures accurately, validate every control, and produce precisely zero risk reduction. Because identifying an exposure and having the authority to make somebody remediate it are entirely different capabilities, and only one of them ships in a product.
The pattern goes like this:
A technically excellent person buys a technically excellent tool.
↓
The tool works.
↓
It surfaces exposures that cross four departments.
↓
And then nothing happens, because the person who bought it doesn't have the standing to make four departments reprioritize their quarters.
↓
Six months later, the tool gets blamed for not delivering value, when what actually failed was organizational authority.
That's not a vendor problem or a buyer problem. It's a structural one, and it shows up everywhere.
We heard a version of the same thing from a completely different vendor two days earlier, framed as three executives with opposed incentives who can never agree on what matters. Same wall, different side of it.
This reframes what a measurement layer is for. A threat debt index isn't really an analytical instrument. It's a political one. Its job is to convert "these forty things are broken" into something an executive can sponsor, defend, and be held to. Findings don't create authority. Numbers sometimes do.
Buying a Capability Versus Running a Program
The related observation is one I'd guess AttackIQ has learned the hard way.
CTEM is not a product you install. It's a program you operate, with defined roles, a cadence, escalation paths, and someone accountable for the loop closing.
And most organizations are not good at standing up new programs, which is worth stating without judgment, because the reason is sympathetic.
Our industry takes excellent engineers and promotes them into management. They know how to operate tools at a level most executives never will. Nobody ever taught them to design an organizational program, define responsibilities across silos, or build the reporting cadence that keeps one alive. Those are genuinely different skills, and expecting them to arrive automatically with a title is a failure of the industry, not of the individual.
AttackIQ's response is to couple professional services with the platform and effectively run the program alongside the customer for the first stretch, on the theory that the only way to teach it is to do it together. Their packaged version of that is a CTEM in 90-day engagement, and the services team is led by John Baker, formerly of MITRE.
I'd flag one more thing they raised that I think is underappreciated across the whole category. Published implementation guidance for CTEM is considerably thinner than the category's prominence would suggest.
There's a great deal of material explaining what CTEM is and why it matters, and much less explaining how to actually stand one up on a Monday morning with the staff you currently have. That gap is why vendors keep having to teach the program alongside the product, and it's a gap the analyst community, ourselves included, should be filling better than we are.
Phase | What buying the capability gives you | What running the program requires |
|---|---|---|
Discovery | Asset and exposure inventory | Agreement on what's in scope and who owns each domain |
Prioritization | Risk scores and attack path graphs | A single accepted ranking that survives four departments arguing |
Validation | Evidence a control works or doesn't | A remediation path with named owners and dates |
Measurement | Dashboards and trend lines | One number leadership will actually be held to |
Sustainment | Product updates and new content | A cadence, an escalation route, and executive sponsorship |
The left column is a procurement exercise. The right column is a change management exercise. Vendors sell the left and get judged on the right.
Where I'd Push
Two things, offered in the spirit of the conversation.
First, the threat debt index needs external scrutiny to become useful. Any single-vendor risk score has an obvious credibility ceiling: it's a number produced by the company selling you the remediation.
The concept deserves better than that, and the route to making it durable is independent methodology review and comparability across vendors. A metric only one company calculates is a feature. A metric an industry accepts is infrastructure. Brett struck me as someone who understands that distinction, which is not universal among people who have just built a proprietary score.
Second, the professional services coupling is both the right answer and a scaling constraint. If the honest requirement is that someone has to run the program alongside the customer for the first ninety days, then growth is bounded by how many people you can hire who are capable of doing that well.
That's a good problem compared with the alternative, and it's still a problem. The interesting question is whether the program knowledge can be productized, templated, or pushed to partners without losing the thing that makes it work.
Neither of these is a criticism of the strategy. They're the two places I'd want to see progress in twelve months.
What To Do Now
For CISOs and security leaders
Pick a number before someone picks one for you. Whether it's threat debt or something you build yourself, arriving at a board meeting with a tracked metric beats arriving with a findings count. The metric doesn't have to be perfect. It has to be consistent and comparable quarter over quarter.
Audit your remediation authority, not just your tooling. For your top ten exposures, write down who has to act and whether you can actually make them. If you can't, more detection capability won't help you.
Separate the tool decision from the program decision. Budget and plan them as two things, because they are. A capability with no program around it becomes shelfware with an annual renewal.
For CIOs and executive sponsors
Ask your security leader what numbers they're measuring. If the answer is a list rather than a metric, that's not a reflection on them. It's a gap in what the organization has asked for and funded.
Understand that your sponsorship is the product. Cross-silo remediation happens when an executive requires it and does not happen when they don't. No platform substitutes for that, at any price.
Fund the program, not only the license. If a vendor tells you their implementation needs services alongside it for the first quarter, that's more honest than the vendors who don't say it.
For practitioners and program owners
If you were promoted for technical excellence, get explicit help with program design. It's a different discipline and nobody expects you to have absorbed it by osmosis. Asking is competence, not weakness.
Insist on an executive sponsor before you buy. If nobody above you is accountable for the outcome, you're buying a tool that will be blamed for an organizational failure.
Key Takeaway
The most useful thing to come out of this conversation had nothing to do with breach and attack simulation.
Security is the only major executive function still operating without an accepted measure of whether it's working. That absence is what produces the accountability model Brett described, where the only feedback signal is a breach and the only consequence is a firing. It's a bad system for the people in it and a worse system for the organizations depending on them.
Fixing it requires two things that don't usually arrive together: a number credible enough that executives will govern against it, and enough organizational authority that the number can actually be moved. AttackIQ is building toward the first and is candid that the second isn't something a product delivers.
That candor is worth more than most product roadmaps. And it's the kind of thing you only hear when the person across the table has done the work themselves.
Working out how to express security program effectiveness in terms your board will govern against, or trying to validate a vendor's risk model independently? Independent research sets the criteria, benchmarks turn cost claims into evidence, and advisory narrows it to your specific decision. GigaOm does all three. Speak with one of our team members today.
Thanks to Brett Galloway for a genuinely unvarnished conversation at Black Hat USA 2026. Views and analysis here are GigaOm's own.
Rob Teel is a Field CTO and Senior Advisor for organizational technology strategy, innovation, and digital infrastructure to drive business growth, competitiveness, and transformational leadership. He is a data, analytics, AI, security, and risk expert in the information technology, government, defense, energy, telecommunications, and healthcare industries. Click here to connect with Rob or schedule an alignment strategy session.
For more, see:
GigaOm Radar for Attack Surface Management
GigaOm Radar for Threat Intelligence Platforms
AttackIQ Breach and Attack Simulation Platform
Breach and attack simulation - AttackIQ