Two Deadlines, One Demand
On 13 July 2026, the Bank of England, PRA and FCA began directly overseeing the technology and service providers that underpin the UK financial system. On 2 August 2026, the transparency obligations of the EU AI Act become enforceable. Two regimes, two jurisdictions, two sets of legal machinery. They make the same demand.
Show us how your processes run.
Not how they were documented three years ago. Not how the policy manual says they should run. How they actually run: where they fail, how they recover, and what decisions get made along the way. And into this demand lands an uncomfortable finding from recent industry commentary: most enterprise AI governance is already out of date.
That is not a coincidence. Out-of-date governance and unevidenced process are the same failure wearing two labels. Both regimes are converging on a single requirement, and most firms are not ready to meet it.
CTP Oversight: Resilience Reaches the Supply Chain
The Critical Third Party regime extends operational resilience scrutiny beyond the regulated firm and into its supply chain. UK regulators now directly assess the technology and service providers underpinning finance, the cloud platforms, data providers, and managed services on which important business services depend. It is the UK’s structural parallel to DORA, which imposes comparable oversight of critical ICT third parties across the EU.
This is a material shift in where the burden of evidence sits. Previously, a regulated firm carried the resilience obligation and managed its providers through contract. Now the provider itself is in scope. Both the firm and its critical providers must be able to demonstrate how key processes run, fail, and recover.
Demonstration is the operative word. A regulator conducting CTP oversight is not asking whether you have a business continuity policy. They are asking whether you can evidence that the process behaves as the policy claims, under stress, in production. That gap between documented intent and executed reality is exactly where operational resilience obligations are most often failed.
Tribal knowledge does not survive this scrutiny. A recovery process that lives in the heads of three senior engineers is not evidenced. It is assumed. When the regulator asks for the record of how a critical service failed over and recovered, the answer cannot be a runbook that may or may not reflect what happened. This is the argument for regulatory defensibility through process intelligence: the ability to show, not assert, how a process actually executed.
EU AI Act Transparency: The 2 August Cliff
Some earlier commentary framed the EU AI Act as delayed, and treated the compliance window as comfortably open. On the transparency obligations, that framing no longer holds. The transparency requirements become enforceable on 2 August 2026. This is not a proposal, a consultation, or a deferred date. It is a live enforcement point.
Precision matters here, because the AI Act’s timeline has several distinct thresholds and they are easy to conflate. The general-purpose AI model obligations run on their own schedule. The transparency obligations are the ones arriving on 2 August, and they carry a specific requirement: firms deploying AI systems in the EU market must be able to explain what a system does, on which process it operates, and with what decision logic.
That requirement is harder than it looks, because AI governance frameworks were mostly written for generative AI, a system that produces content on request. The frontier has moved to agentic AI, where systems take actions inside live business processes. A governance framework that describes prompt controls and content filtering says nothing about which process an agent is executing, where it holds authority to act, and where a human decision is required. This is why so much enterprise AI governance is already out of date: it governs the wrong layer.
Transparency, in the sense the AI Act now enforces, is a process claim before it is a model claim. You cannot explain what a system does without first being able to articulate the process it operates on. This is process articulation as AI Act evidence: the AI Act’s transparency obligation resolves, in practice, to a question about whether you can see your own processes clearly enough to explain them.
The Common Substrate: Evidenced, Articulated Process
Strip both regimes back to their operative demand and they reduce to the same question. Can you evidence how the process runs?
CTP oversight asks it of your critical services under stress. AI Act transparency asks it of the processes your AI systems touch. In both cases, the failure mode is identical: an organisation that can describe what should happen but cannot evidence what does. Out-of-date governance is not primarily a documentation problem or a legal problem. It is a visibility problem. You cannot govern, and you cannot evidence, what you cannot see.
This is the substrate IGX360 Insights is built to provide. Its resilience lens shows how a process behaves under failure and recovery, the exact evidence CTP oversight demands. Its provenance and audit-readiness turn a live process into a tamper-evident record of what happened, when, and under whose authority. The process stops being a diagram that ages on a shared drive and becomes a governance asset that reflects execution.
Articulated, evidenced process is not a separate answer to each regime. It is the single answer to both. Build it once, and CTP oversight becomes demonstrable and AI Act transparency becomes defensible from the same foundation. That is the point of treating process as audit-ready process insights rather than as static documentation.
Get Your Processes on Record Before 2 August
The transparency enforcement point is a fixed date, and it is close. CTP oversight is already live. The readiness window for both is not opening, it is closing.
The work is not to write more policy. It is to make your processes visible and evidenced, so that when a regulator asks how a critical service recovers or what an AI system does, you can show the record rather than describe the intention. IGX360 Insights is built for risk leaders and compliance teams facing exactly this demand.
Two regimes are now asking the same question. Can you show them, or can you only tell them?
Turn your processes into audit-ready evidence: request a diagnostic.