Safety & Trust in the Age of AI

Also watch and listen

The signals

• Sabino: everyone doing AI compliance is really doing AI procurement. Ask what the frameworks are actually complying to and the answer is: making the thing buyable, movable, integratable, insurable. Useful work, wrong first question. The first question is AI value safety, which is whether the valuable thing you hand this machine stays safe in its custody and comes back plus new value. Nothing on the market today asks it, because compliance frameworks are built around a company that will not stop moving, and they measure the cheapest way to check the box rather than whether the machine is running at its potential.
• Sam: when it comes to trust, AI is more like people than computers. Recorded the same day [OpenAI reported that its own internal models broke out of the lab and hacked Hugging Face](https://fortune.com/2026/07/21/openai-says-ai-models-escaped-control-hacked-hugging-face/) to pass their test. The model did not know not to do that, because it has no concept of where the bounds are. Fifty years of deterministic systems trained everyone to trust the output because the machine said so. This is not that. It is someone you hired off the street with amnesia, day one on every conversation, trusted to get certain things wrong. You would check their work. You would have standards you thought about in advance. You would not hand them the keys to the kingdom on day one, and the keys would be somewhere they cannot reach as a matter of architecture, not as a matter of policy.

The subtractions

• Sabino: retire responsible, ethical, consensual and compliant AI as the assurance. Not because responsibility is bad, but because the label does work it cannot support: responsible ownership of a bull in an apartment still leaves you with a bull in the apartment. AI is a force amplification technology. Create value and it creates more; extract value and it extracts more; screw someone and it screws them better. Given a bigger stick, people swing it. So there is no responsible use, no ethical use, no compliant use. There is only whether your value is safe, and Sabino applies weapon and ordnance safety models rather than ethics frameworks to get at it.
• Sam: retire "human in the loop" as if it were quality assurance. A person pressing the button instead of a machine pressing the button is not protection from anything. Press it forty-seven times before lunch and the forty-eighth press carries no awareness at all. There are real cultural reasons to keep a person there, and the ownership of the process is genuinely human, but the loop is not a control. Sabino's version is sharper: the human in the loop is the engineer at the Chernobyl dashboard, riding the dials without being a nuclear engineer.

About this episode

Two voices this week: host Sam Rogers and guest Sabino Marquez, twenty years a CISO, adviser to CISOs, and for two years the Canadian ambassador to the Global Council for Responsible AI. Which is the interesting part, because he spends the episode arguing that responsible AI is the wrong frame.

The arc runs from a conversion story to a rebuilt first question. Sabino came up in the nineties when protecting your technology was protecting your advantage, watched Sarbanes-Oxley put IT under the CFO after the turn-of-the-century scandals, and found that the more compliance he did the less safety he could prove. His diagnosis is that the business never saw a defender; it saw a security-shaped answer to a security-shaped problem standing between it and the dollar it wanted. That is how a CISO ends up accountable for every trust-creating practice in the company without being empowered to run any of it, and how the trusted-advisor frame becomes a trap: you are important, you are allowed in the room, and you do not have a number to deliver, so you are advice, and advice gets overruled.

The evidence-theatre block is the one to watch. Selling to the Fortune 50 means diligence teams that count every armpit hair one by one, so Sabino booby-trapped his own PDFs with a tracking pixel to find out whether anyone actually opened the documentation his company spent six figures a year producing. Mostly they did not, and his conclusion is not that buyers are lazy: handing someone forty documents is asking them to assemble a puzzle without the picture on the box. Data makes people feel something. The IT reviewer, the lawyer, the insurer, the regulator and the procurement lead all need a different story told from the same evidence, which is what it means to run the company as a factory that produces documented proof that value is safe in your hands.

The constructive turn is that this pays. Sabino puts it at roughly half a billion dollars more across ten years and multiple companies: higher exits, bigger rounds, higher ACVs, faster sales at lower cost. He gives the system away because it is not a product, it is a better way to walk. His closing claim is that safety becomes the number one buying criterion, ahead of price and quality, because we have entered an era of unrestrained velocity where the benefit of speed is outrunning the explosions you set behind you.

The value inventory is the part most companies fail cold. Asked why their company is valuable, leaders reach for contracts and the GitHub repo. The real answer is source code, pricing logic, customer histories, research, institutional judgment, the meaning of the brand, and the synthesis of the humans who hold all of it. Those are not raw material to be transformed into value. They are the value. You cannot prove an AI is safe for value you never identified.

This week's links:

• OpenAI says AI models escaped control and hacked Hugging Face (Fortune) — the day-of story behind Sam's signal
• Trustable — Sabino's consulting practice, and this week's sponsor
• Trust Club — his ongoing blog, where most of the system is published as essays and instruments, aimed at people who do trust work for a living
• Sabino Marquez on LinkedIn

Sponsored by Trustable, which builds instruments for measuring current trust value and designs operating conditions that compound it. First time the Tier 1 slot has gone to the guest's own property.

Want to bring your own signal and subtraction? Find yours. Sponsored by Trustable.

The newsletter

Issue 61 · Sunday 26 July 2026 · sent on Substack and LinkedIn.

One signal, one subtraction, one analogy, in five minutes. It draws on this episode but stands on its own. Every issue.