Every design principle you were trained on told you to remove friction. The FTC just fined Epic Games $520 million for following that advice too well.
Seamless is not the goal anymore. Intentional is. And the field is only beginning to understand the difference.
The Principle That Became a Weapon
Remove friction. Streamline the flow. Eliminate barriers between the user and the action. These are not fringe ideas. They are the foundational vocabulary of user experience design, repeated in every UX education, every design review, and every product brief the field has produced for thirty years.
They are also, applied without judgment, one of the most reliable mechanisms the industry has ever built for producing harm at scale.
The FTC’s Click-to-Cancel Rule, enforced as of 2025, was a direct legislative response to a design philosophy that had been used deliberately against users for years. Companies made joining a service a single click and leaving it a multi-screen ordeal, an approach the FTC named the Iliad Flow, a reference to a journey so long and arduous it was borrowed from a nine-thousand-line epic poem. Epic Games’ $520 million settlement emerged from interface design that made it easy for users, including children, to trigger high-value purchases by reflex or subtle visual misdirection. The logic that produced these outcomes was not the logic of malicious actors. It was the logic of designers optimizing for conversion metrics using the same frictionless philosophy the field had been building on for a decade.
The 2026 rule that emerged from this reckoning is precise: if a user can join in two clicks, they must be able to leave in two clicks. That regulation is not a user experience improvement. It is a correction imposed on an industry that had turned its own defining principle into an extraction mechanism, and it signals a deeper shift in what good UX design is now understood to require.
What Every UX Boom Got Wrong About Ease
Every UX paradigm in the field’s history has produced a version of the frictionless ideal, and every version has created the conditions for its own abuse.
The GUI era made computing accessible by removing the cognitive friction of command-line interfaces. This was a genuine and important contribution to human-computer interaction. It was also the first step toward an interaction model where the easier something was to do, the less the user had to think about whether they actually wanted to do it. The ability to double-click a file and watch it open without understanding what was happening underneath it was the beginning of a design culture that traded user understanding for user convenience, a trade whose costs only become visible when the convenience is applied to consequential decisions.
The mobile era intensified this trade. One-tap purchases, saved payment credentials, default opt-ins, auto-renewal subscriptions: each of these reduced friction in a way that served users when they were aware of what they were agreeing to and systematically disadvantaged them when they were not. The mobile design vocabulary that emerged from this era, streamlined onboarding, reduced step counts, minimized confirmation requirements, was taught and applied as though frictionlessness were a user benefit independent of context. It was not. It was a design choice with a specific cost-benefit profile that varied dramatically depending on the stakes of the action being streamlined.
The conversational UI era made frictionlessness conversational. A single sentence could authorize an action, place an order, or initiate a process with no visual review of what was being agreed to. The design challenge of making consequential actions legible when the interface was language rather than a screen was largely unresolved because the field had not yet developed the vocabulary for friction that serves users. It had only the vocabulary for friction that impedes them.
Research published in 2025 on AI-supported decision-making named the systemic consequence of this trajectory: automation bias, where users uncritically accept AI recommendations, degrades human skills and weakens professional accountability. A field that spent thirty years removing friction from human decision-making has been systematically reducing the deliberateness with which people make consequential choices, and the AI era has simply made the consequences of that reduction more rapid, more powerful, and more difficult to reverse.
Why the Ambient Intelligence Era Makes Friction Design the Most Important Skill in UX
The next wave of UX is driven by ambient intelligence, emotional context, and zero-UI experiences. Each of these forces operates in a domain where the cost of frictionless design applied without judgment is higher than it has ever been, and where intentional friction is the design intervention that separates systems that serve users from systems that capture them.
Ambient intelligence means systems that act on the user’s behalf without being summoned. A system with no designed friction between observation and action is a system that can make consequential decisions at the speed of its inference, without any mechanism for the user to reflect, reconsider, or understand what is being decided. Research on AI-supported decision-making published at SAGE journals in 2025 proposes designed friction as intentional design interventions that introduce interaction friction at critical points as a mitigating factor to inappropriate reliance, arguing that AI embedded with meaningful friction can foster reflection, preserve human agency, and promote appropriate trust. The key phrase is at critical points. Friction is not appropriate everywhere. It is essential somewhere, and designing for which somewhere requires a level of judgment the field has not yet formally developed.
Emotional context makes this more urgent, not less. A system that reads a user’s emotional state and adjusts its behavior accordingly is a system that has unique insight into when a user is most vulnerable to impulsive, regrettable, or manipulated action. A frictionless ambient system that reads high emotional arousal and surfaces a high-stakes decision in that moment, without any designed pause, is not serving the user’s long-term interest. It is serving the product’s conversion metric. The design of emotional-state-aware friction, the system that detects vulnerability and adds a pause rather than removing one, is one of the most ethically demanding design challenges the ambient intelligence era has produced.
Zero-UI removes the visual confirmation layer that previously provided at least a minimal form of friction for most user actions. When there is no screen, there is no moment where the user sees a summary of what is about to happen before it happens. The action occurs. The outcome follows. Designing confirmation and reflection mechanisms for invisible systems requires an entirely new interaction vocabulary, and it requires a design team that has already developed the philosophical commitment to friction as a tool rather than an obstacle.
The Three Shifts That Define Intentional Friction Done Right
Shift 01: Calibrate friction to the reversibility of the action, not to the conversion goal
The principle that should govern every friction decision is not how do we make this faster? but how much reflection does this action require, given what happens if the user gets it wrong and wants to undo it? GitHub’s repository deletion flow requires users to type the exact name of the repository to confirm. AWS console applies similar confirmation requirements to one-way-door decisions like deleting endpoints or EC2 instances. These are not bad UX decisions penalizing users for trying to accomplish something. They are appropriate friction applied to irreversible actions that have serious consequences if performed by reflex or mistake. The design principle is proportionality: friction that matches the stakes and reversibility of the action it is protecting. A low-stakes, easily reversible action should be as frictionless as the design can make it. A high-stakes, irreversible action should have friction that is proportional to what the user is agreeing to give up if they proceed.
Shift 02: Design friction that informs rather than friction that frustrates
Research on security-enhancing friction establishes the critical distinction between friction that helps users make better decisions and friction that simply makes tasks harder without producing any benefit to the user’s understanding or protection. Design friction refers to elements designed to provide difficulty during an interaction with technology, and when friction is designed with intention, the interaction can cause reflective, informative, and safe behavior among users while improving user experience. The difference is in the quality of the friction: does the pause, the confirmation, the explanation, the delay serve the user by giving them information they need to make a better decision? Or does it serve the product by adding a step that the product can point to as a safety measure while producing no actual improvement in user understanding or protection? Friction that informs is a design contribution. Friction that frustrates is a UX liability that will eventually be regulated away, as the FTC’s Click-to-Cancel Rule demonstrated.
Shift 03: Build the anti-frictionless audit into every design review
The design review process that most product organizations run asks, consistently and often only, whether the experience is easy enough. The question that needs equal standing in every design review is whether the experience is deliberate enough. For every consequential action in a product, the design team should be able to answer: what prevents a user from triggering this action by reflex, by mistake, by emotional impulsivity, or by subtle interface misdirection? What does the user understand about the consequences of this action before they confirm it? Is the confirmation mechanism proportional to the stakes? If the action is made easier for the user to trigger, does that benefit the user or the product? These questions do not slow design down. They redirect it. And in an era where regulators are now writing the answers into law, getting there voluntarily produces better products and avoids the alternative, which is getting there by settlement.
The Closing That Should Show Up in Your Next Sprint
Here is the honest assessment of where the field is on friction design right now.
Research published in 2026 identified what it called an agentic takeover in the trajectory of UX research: a brief surge in 2025 defending human epistemic sovereignty, meaning the user’s right to think and decide for themselves, was suppressed by an explosive shift toward optimizing autonomous machine agents, while frictionless usability maintained what the researchers called structural hegemony at 67.3 percent of the design conversation. The field is moving faster toward autonomous systems that act on users’ behalf than toward the friction design vocabulary that would make those systems trustworthy.
The regulation is already responding. Security and UX design are no longer separate concerns, with growing regulatory requirements around informed consent and data protection. The FTC’s Click-to-Cancel enforcement is active. Epic Games’ $520 million settlement is on the record. The EU’s work on AI transparency and explainability is establishing friction requirements for AI-influenced interfaces across member markets. The regulatory direction is not ambiguous. Designers who build the intentional friction vocabulary now are ahead of the compliance requirements that are being written around them.
Making products harder to use in specifically chosen moments enhances UX security and user trust. That sentence, from research on security and UX design, is the clearest formulation of what intentional friction design actually is: not a failure of usability but a fulfillment of it, in the moments where usability and safety have been placed in false opposition by a design culture that confused ease with goodness.
Seamless is not the destination. Intentional is. And the designer who can tell the difference between friction that serves the user and friction that serves the product is the designer who will build the products that both users and regulators trust.
Build the pause where it matters. Remove it everywhere else. That is the practice.
Research sources: SAGE Journals, Better AI with Designed Friction: Theories, Applications and Research Agenda, 2025; Medium Design Bootcamp, The Friction Manifesto: Why Seamless Is the New Dark Pattern, April 2026; ACM Digital Library, The Framework of Security-Enhancing Friction, 2021; Toptal UX, Necessary Friction: The Theatrics of UX Security, 2024; Ping Identity, Security by Design: Building Trust Without Sacrificing UX, March 2026; UXMatters, Secure UX: Building Cybersecurity and Privacy into the UX Lifecycle, March 2025; ResearchGate, The Impact of Security-Enhancing Friction on UX; arxiv, Cognitive Agency Surrender: Defending Epistemic Sovereignty via Scaffolded AI Friction, 2026; FTC Click-to-Cancel Rule enforcement documentation 2025; Epic Games FTC settlement reporting 2022-2026.