The obligation to evidence communication outcomes | Digisarion Intelligence
Consumer Duty board reporting, EU AI Act Article 50 and the Online Safety Act have made communication outcomes an evidenced obligation. What each requires, and who is personally accountable.
The obligation is dated, recurring, and signed by a named person.
Communication outcomes used to be a matter of good practice. Across three separate regimes, they are now an evidenced obligation with deadlines attached and an individual accountable for meeting them.
Consumer Duty
FCA Consumer Duty: the consumer understanding outcome
The Consumer Duty requires firms to support consumer understanding, so that customers receive communications they can understand and are equipped to make decisions that are in their interests.
The requirement that bites hardest is not the standard itself but the reporting. Firms must produce an annual board report assessing whether they are delivering good outcomes, and communications sit inside that assessment. It is a recurring, dated obligation with a named owner, which is why it behaves like a purchase trigger rather than an improvement project.
The regulator has also been explicit that firms unable to demonstrate differentiated consideration for customers in vulnerable circumstances are not meeting the standard. That raises the evidential bar considerably: it is no longer sufficient to show that a communication was clear on average.
EU AI Act
EU AI Act Article 50: transparency obligations from 2 August 2026
Article 50 imposes transparency duties on providers and deployers of certain AI systems, including obligations around content that is artificially generated or manipulated. The obligations apply from 2 August 2026.
For any organisation generating customer-facing communications with AI, this converts a design choice into a compliance requirement, and it applies irrespective of whether the organisation considers itself an AI company.
Penalties under the Act reach the greater of 35 million euros or seven per cent of total worldwide annual turnover, depending on the infringement.
Online Safety Act
UK Online Safety Act: duties with an evidence requirement
The Act places statutory duties on in-scope services, with Ofcom empowered to require information and to assess whether duties are being met. As with the Consumer Duty, the practical burden falls on being able to evidence the position rather than simply assert it.
GEI-AI is in development. The platform is at an early research and build stage. No product is commercially available, no deployment has taken place, and we are not making any claims about its current capabilities.
Who carries it
The duty attaches to a person, not a department.
Under the Senior Managers and Certification Regime, accountability for these outcomes rests with a named senior individual, typically the Chief Compliance Officer or Chief Risk Officer. They sign the assessment. They answer for it if it is wrong.
That changes the buying question. It is no longer "would this improve our communications", which is a budget conversation. It is "can I evidence this to the board and to a regulator", which is a personal one.
Why the current method struggles
Sampling tells you about the sample.
The prevailing approach is manual review of a sample of communications, scored against an internal rubric by trained reviewers.
Three difficulties follow. Coverage cannot be claimed, because the population is far larger than any sample a team can read. Consistency is hard to demonstrate, because two reviewers assessing the same document frequently disagree and the disagreement is rarely measured. And selection works against you: the communications most likely to fail, such as arrears correspondence, complaint responses and adverse decisions, are typically written under pressure and are not the ones a routine sample tends to surface.
None of this is a criticism of compliance teams. It is a description of what happens when a duty framed around an entire population is discharged with a method that can only address part of it.
What we are building against this.
GEI-AI is being built to score communications against these outcomes at population scale, and to produce documentation a board and a regulator can read. It is in development, and no product is currently available.
We are an assurance tool. We strengthen the evidence a firm produces about its own communications. We are not an independent third-party audit, and we do not certify compliance.