How should loyalty marketing professionals interpret the Citrini Research report?
The viral “2028 Global Intelligence Crisis” essay from Citrini Research and Alap Shah is explicitly framed as a scenario, not a forecast. That distinction matters, as the authors are not claiming certainty as much as they are testing a plausible downside case in which AI adoption accelerates faster than labor markets, institutions, and business models can adapt.
For leaders in loyalty, CRM, and gamification, the essay’s real value is not its macro drama (of which there has been plenty, much to the consternation of Wall Street). I see it more as a pressure test to a much more immediate “read-between-the-lines” question: what happens when AI strips away the friction, confusion, and inertia that many customer strategies quietly rely on? That question cuts to the center of modern loyalty design, and the essay offers some compelling, and even frightening, scenarios to consider.
Many programs are presented as engines of retention and brand affinity, but too many still depend on a weaker formula: opaque points economics, promotional habit loops, and switching costs created by inconvenience rather than genuine preference. Those mechanics have always been fragile. AI simply makes the fragility harder to hide.
If consumers increasingly use AI agents to compare offers, manage subscriptions, optimize spending, and route purchases toward the best available value, then “good enough” loyalty structures begin to fail faster. Programs that work because customers are busy, distracted, or unwilling to do the math may not merely lose efficiency. They may lose their advantage altogether.
This is the loyalty industry’s version of a ghost metric problem. A program can produce impressive activity data and still fail to create durable preference. It can generate opens, redemptions, repeat transactions, and app sessions while doing very little to increase the likelihood that a customer will choose the brand when a cheaper or easier option appears. In that sense, many brands have been measuring motion and calling it loyalty.
AI intensifies that gap because optimization is exactly what it does best. If a customer-side agent can instantly evaluate price, terms, fees, redemption value, and alternatives, then weak loyalty design gets stress-tested continuously. What survives that environment is not gimmickry or opacity, but clear value and real relationship.
That is also why AI will likely split the gamification market in two. Superficial tactics (see: badges layered onto discounts, “missions” that are just coupons in costume, engagement mechanics built to inflate dashboards) will become easier to exploit and easier to ignore. By contrast, stronger “brand playability” may become more valuable. Systems that create progression, mastery, anticipation, identity, and belonging operate on a different level than transactional optimization. AI can optimize a purchase. It cannot replace the human desire for achievement, recognition, or meaning.
The implication for executives is straightforward. The future is not a choice between machine efficiency and human experience. Brands will need both. They will need machine-readable value that stands up to AI scrutiny, and they will need human-centered relationship design that gives customers a reason to prefer them beyond price. If they build only the first, they risk commoditization. If they build only the second, they risk being bypassed at the moment of transaction.
A useful leadership test is simple: if an AI agent managed your customers’ purchases tomorrow, why would your brand still win? If the answer is mainly points, promotions, or habit, the strategy is vulnerable. If the answer includes trust, utility, emotional memory, and a relationship customers actively want, the foundation is stronger.
The Citrini essay may or may not prove directionally correct in every macro detail. But for loyalty and gamification, it already surfaces a truth the industry has postponed for too long. As intelligence becomes abundant, weak loyalty strategies stop hiding. That is not just a threat. It is a long-overdue audit.