16/09/2026
Can a model that operates your computer for you also decide what "AI-generated" means for your business?
OpenAI has released GPT-6 Astra, a model built to browse, code, fill in forms, manage records and carry out full pieces of professional work with very little human input along the way. It can sit inside your operations and act on your behalf, which means the question of what it is allowed to do matters more than it did last month.
Many DMOs carry a quiet worry about this. A full-scale AI rollout feels risky when it touches people's roles and livelihoods, and that hesitation is a reasonable starting point. It is also exactly why a framework matters most. A framework lets you adopt AI on your own terms, from broad use down to none at all, and gives your team clear limits so everyone knows what is allowed and where the line should be drawn. Choosing to stay away from AI altogether is a valid position too, as long as it is written down and shared clearly, so every member of staff is working from the same understanding and shadow AI has nowhere to take hold.
Setting those limits well means being specific and that is where a lot of early frameworks fall short. Version 1 of our AI Transparency Framework asked a single question: how much of a piece of work was AI-generated, on a percentage from 1 to 99. Testing that question with our DMO peer group surfaced two problems. A percentage implies a precision nobody could honestly give and that sat uneasily in a framework built on honest disclosure. It also treated a strategy document and a generated image as the same kind of work, when AI supports each of those in different ways.
Version 2.0 answers both of those points. You start by choosing the medium from 9 options, including an article, a report, an image, a social post, a video, a website, an email, a presentation or a bid. Each medium carries its own set of activities and those answers are weighted and combined into a score. This gives DMOs specific, workable limits to build a policy around.
π Explore Version 2.0 of our AI Transparency Framework: https://ai.thinkdigital.travel/