September 2026

The Regulators Have Caught Up: What Dark-Pattern Enforcement Means for Your Site

Posted on September 23, 2026September 22, 2026 by administir

dark pattern enforcement, image of confused navigation

Key Takeaways

“Dark patterns” (interfaces designed to nudge users into decisions they wouldn’t otherwise make) are no longer a purely ethical concern. They’re a regulatory one.
The FTC’s click-to-cancel rule and the EU’s Digital Services Act now impose real penalties on flows that make cancellation, opt-out, or consent unnecessarily difficult.
Common offenders include: subscription cancellation flows with more steps than sign-up, pre-checked consent boxes, roach-motel account deletion paths, and misleading urgency cues.
Fixing these before enforcement arrives is meaningfully cheaper than fixing them after. And the sites that fix them tend to convert better, not worse, once the manipulative friction is removed.
The test worth applying: would a regulator reading your flow the way a skeptical user would find anything to object to?

A few years ago, “dark patterns” was the sort of phrase you’d hear at UX conferences and industry meetups. Designers and usability practitioners argued about it in blog posts. Ethical brands made noises about avoiding them. Meanwhile, plenty of large consumer sites kept using them, because they worked, and because there was no real cost to using them.

That cost has now arrived. Over the last eighteen months, regulators on both sides of the Atlantic have moved from strongly-worded guidance to actual enforcement, and the design decisions that used to be a matter of company values are now a matter of company liability.

This post is a quick look at what changed, what’s now genuinely risky, and what to do about it if you haven’t looked at your own site’s flows lately.

What actually changed

Two regulatory moves are worth knowing about specifically.

The FTC’s “click-to-cancel” rule, finalized in late 2024 and now in active enforcement, requires that cancelling a subscription must be at least as easy as signing up for one. If a customer can subscribe with two clicks online, they must be able to cancel with two clicks online. No mandatory phone calls, no navigating through five screens of retention offers, no chat queues designed to time-out inactive users.

The EU’s Digital Services Act (DSA), fully in force across the region, prohibits design that “deceives or manipulates” users, with explicit callouts for pre-checked consent boxes, misleading prominence of options, and repeat prompting for choices the user has already made. Penalties can reach 6% of global annual turnover for large platforms.

Neither regulation is aimed solely at giants. Both apply to consumer-facing sites that serve users in those jurisdictions, and both have been used against smaller operators as test cases. If you have US or EU customers and you run subscription, consent, or account flows on your site, these apply to you.

The offenders you’re most likely to have

Most sites don’t set out to be manipulative. The dark patterns creep in gradually, usually as small optimizations that each seemed reasonable at the time. The most common ones seen in audits include:

Cancel flows longer than sign-up flows. The classic example. Customer signs up in three clicks; cancelling takes seven screens, three retention offers, and a phone call. Now explicitly regulated in the US.
Pre-checked consent boxes. Boxes for marketing consent, data sharing, or newsletter opt-in that are checked by default. Explicitly prohibited under the DSA and under most modern privacy regimes; also poor practice under longstanding US guidance.
Roach-motel account deletion. Making it easy to create an account and effectively impossible to delete one. Common on older platforms; increasingly a source of complaints and regulator interest.
Confirmshaming. Opt-out language written to make users feel bad for declining. “No thanks, I don’t want to save money.” While it may read as clever copywriting internally, it reads as manipulation to a regulator.
False urgency cues. Fake countdown timers, fabricated stock scarcity (“only 2 left!” when there are 200), or fake activity indicators (“3 people are viewing this now”). Increasingly scrutinized by both the FTC and EU regulators.
Cookie banners that make “accept all” one click and “reject all” three. Any consent flow where the manipulated choice is easier than the neutral one now falls squarely inside the DSA’s definition of manipulation.

Why fixing this is good business, not just legal hygiene

The counterintuitive thing that keeps coming out of site audits is that the sites that clean these patterns up tend to convert better afterwards, not worse.

Manipulative friction produces short-term compliance and long-term resentment. Users who feel coerced into subscribing cancel harder when they eventually do; users who feel tricked into consent are the ones who leave one-star reviews and file complaints. Users who found the flow honest and easy are the ones who convert, stay, and refer.

The economics of dark patterns almost always look worse than they seem once you factor in the churn, the support tickets, the chargebacks, and the reputational drag. Fixing them isn’t a favor to your users; it’s a favor to your overall business health.

What to actually do

If you haven’t looked at your own site’s flows through this lens recently, three practical things worth doing:

Do a cancellation audit. Count the clicks from your sign-up flow. Count the clicks from your cancellation flow. If cancellation is longer, that’s your fix.
Do a consent audit. Check every place your site asks for consent (cookies, marketing opt-in, data sharing, newsletter). If “accept” and “decline” aren’t equally easy, they need to be. If any boxes are pre-checked, they need to not be.
Do a skeptical read. Have someone who doesn’t work for the company walk through your site’s most critical flows (sign-up, checkout, cancellation, account settings) with a single question: is anything here trying to nudge me toward a choice I wouldn’t otherwise make? Anywhere the answer is yes deserves a second look.

The regulatory environment on this is unlikely to loosen. If anything, enforcement will get stricter over the next few years as regulators build case law and audit patterns. Sites that address these patterns before enforcement arrives will do so on their own timeline and their own terms. Sites that wait will address them under external pressure, which tends to be more expensive and less thorough.

It’s a good time to look at your flows. Better now than after a complaint.

About the author

Neil Fraser leads digital strategy at STIR, helping clients translate brand positioning into websites and digital experiences that actually convert. He works across discovery, UX, and build, with a particular focus on the details that quietly determine whether a site earns its keep, and, increasingly, whether it stays on the right side of regulators.

How AI Delivers More Transparent Account Management

Posted on September 8, 2026September 8, 2026 by administir

Key Takeaways

The client-agency status document is doing more work than most people realize, and doing it poorly. Every status meeting starts with someone stating what happened last week from memory and scrolling through emails.
AI has meaningfully shifted what account management can actually deliver: real-time visibility into where projects stand, cleaner meeting recaps, faster recap drafts, better project reviews.
This isn’t about replacing account managers with AI. It’s about giving account managers the tools to do the accountability work that’s always mattered but rarely got done well.
For clients, the shift shows up as fewer surprises, faster answers, and status updates that reflect reality.
The agencies that adopt these tools thoughtfully will reset what “in the loop” means. The ones that don’t will keep leaving their clients guessing.

The status document has to go.

Every client-agency relationship has one. Sometimes it’s a Google Sheet, sometimes it’s a shared doc, sometimes it’s a slide in the recurring meeting deck. Whatever form it takes, it’s supposed to answer the question “where do things stand?” and it’s supposed to answer it clearly enough that both sides can walk into the meeting knowing what’s what.

It rarely does. The status document is usually a rough recap of the last week, thrown together by whoever had five minutes on Monday morning, and it typically reflects what the account manager remembered while updating it rather than what actually happened. The meeting then spends its first fifteen minutes filling in the gaps. Which is fine, except that’s fifteen minutes on the wrong initiative.

This isn’t a criticism of account managers, who are typically doing their best inside a job that has too many moving parts to keep track of manually. It’s a criticism of the tools we’ve been using to do the work, and it’s the reason the shift happening right now matters.

What AI actually changes about account management

This isn’t about AI replacing the account manager. AI is genuinely bad at the parts of account management that matter most: reading the room, picking up on what’s not being said, knowing when to push and when to hold off. What AI is genuinely good at is the parts that most account managers privately dislike: the “paperwork” after the meeting is over, the status gathering, the recap writing, the notetaking, the trying-to-remember-what-was-said work.

A few places we’re already putting this to work:

Meeting recaps that reflect what actually happened. AI transcription and summary tools now produce recap notes that are meaningfully better than what most humans write. The account manager reviews and refines rather than piecing it together from memory. Clients get recaps faster and with more precise detail.
Status documents that self-populate from project management tools. Instead of the Monday-morning scramble, the status dashboard pulls current state from where the work is saved (project management tools, email threads, cloud file storage) and surfaces what’s changed. The account manager verifies and adds context rather than starts from a blank page.
Real-time visibility, no email required. Clients don’t have to send a “what’s the status of X?” email and wait for the account manager to ping three internal team members for an answer. They can open the dashboard and see exactly where a deliverable stands, who has it, and what’s next. That’s not just an efficiency gain; it’s the difference between an agency that feels transparent and one that feels closed off.
Better campaign and project recaps. The end-of-campaign recap is where the best learning happens, and it’s also the meeting that often gets skipped. AI takes the heavy lifting out of pulling the campaign recap together, which means it actually happens instead of getting pushed to “next week” until it never happens at all.

Why this is a transparency story, not just an efficiency story

The efficiency gains are real, but they’re not the interesting part. The interesting part is what the efficiency unlocks.

When account managers spend less time manually updating status docs, they spend more time on the things that actually shape the client relationship: analyzing the account, spotting potential issues before they become problems, connecting dots between projects that would otherwise sit in silos, and having conversations about strategic direction rather than tactical execution. That’s where account management earns its keep, and it’s where most account managers have always wanted to focus but are stuck in the day-to-day tactics.

For clients, this shows up as a genuine change in what the relationship feels like. Fewer surprises. Faster answers. Status updates that reflect reality rather than a best guess at what happened last week. Meetings that spend their time on the actual work rather than catching everyone up on where things stand. That’s what transparency in an agency relationship should mean, and it’s what a lot of clients haven’t reliably had.

What we’ve kept the same

A few things haven’t changed and shouldn’t. The account manager is still the face of the relationship. The strategic thinking, the judgment calls, the read on where an account actually is: that’s still us. When to push, when to pull back, how to handle a difficult conversation, how to frame hard news to a client. Those are human decisions and always will be. AI does the work behind the thinking. It pulls the notes, drafts the recap, surfaces what’s changed. We decide what it means, what to do about it, and how to say it.

Confidential client information is still handled with the same discipline it always has been. AI tools that touch client data are vetted for security and privacy compliance the same way any other vendor would be. The efficiency isn’t worth much if it introduces new risks around confidentiality.

And critically, the relationship still lives in conversations, not in tools. The best account managers use AI to make the tools invisible so that more of the relationship can happen in the actual conversations. Clients don’t want a portal. They want a partner.

Most agency relationships still run on Monday-morning guesswork and status meetings that spend half their time figuring out what happened last week. Yours doesn’t have to. If you want an agency that shows up already knowing where things stand and where they can go, let’s talk.

About the author

Alli Hughes is Account Management Director at STIR Advertising & Integrated Messaging. With more than a decade of experience in marketing strategy, content, and new business development, she leads STIR’s client relationships across brand strategy, digital, and integrated engagements. At STIR, she works directly with clients to build integrated marketing strategies that drive meaningful growth.