DraftKings’ AI Targeting Scandal: 7 Ethical AI Marketing Lessons for Your Business
On September 19, 2026, The New York Times published an investigation that should make every business owner uncomfortable. According to the Times, DraftKings — one of America’s biggest online gambling companies — built a machine-learning system in 2023 that ranked its online casino players by how much money they were expected to lose after receiving a free bet or bonus. Players with the highest expected losses were then flooded with promotional incentives to keep them wagering.
Let that sink in. Not the players most likely to enjoy the platform. Not the most loyal customers. The ones most likely to lose.
Whether you run a local service business, an e-commerce store, or a national brand, this story is about you. Because the same AI targeting technology DraftKings allegedly used is available to every advertiser on Meta, Google, and TikTok today. The question is no longer whether you use AI in your marketing — it’s how you use it, and where you draw the line. Ethical AI marketing is no longer optional; it is a competitive advantage. Businesses that commit to ethical AI marketing now will own the trust advantage for the next decade.
Here are seven lessons every business should take from this scandal.
1. Personalization without a conscience is just manipulation at scale
Here’s how the system reportedly worked. DraftKings’ model analyzed each player’s betting frequency, daily account balances, loss-to-wager ratios, and even the probability that the player was about to quit the platform. It combined these into an internal figure employees called an “elasticity” score. A higher score meant the customer was worth chasing with more perks — more free bets, more bonuses, more reasons to keep playing.
A former DraftKings data analyst who tested the system, Jayden Butts, told the Times exactly what the math was optimizing for: “We are looking for traits and features that we can target that indicate a good investment. The best investment would be a problem gambler.”
Read that quote twice. This is what happens when an AI system is given a revenue target and no ethical boundaries — the opposite of ethical AI marketing. The algorithm didn’t malfunction — it worked exactly as designed. It found the most profitable audience with ruthless efficiency.
The lesson: AI doesn’t have values; it has objectives. If your marketing objective is purely “maximize revenue per user,” your AI tools will happily find the most vulnerable version of your audience to do it. Ethical AI marketing starts with constraining what the system is allowed to optimize for — not just celebrating how well it optimizes. Make ethical AI marketing the constraint, and optimization becomes an asset instead of a liability.
2. Your data can protect people or exploit them — the choice is yours
Perhaps the most damning detail in the investigation isn’t what DraftKings built. It’s what they didn’t build.
According to the Times, a DraftKings data scientist named Nestor Hernandez began building a model in 2024 to flag gamblers sliding toward crisis — essentially the same technology pointed in the opposite direction, toward protection instead of profit. The project was shelved. Chief Responsible Gaming Officer Lori Kalani told the Times that leaders reached a “collective decision” against predictive tools because the approach was not “evidence-based.”
So the company had the data, the talent, and the infrastructure to identify at-risk customers. It used that capability to target them with promotions instead.
The lesson: ethical AI marketing starts with a choice — most businesses today sit on customer data that could be used either way. Purchase histories, engagement patterns, support tickets — these can power genuinely helpful personalization or they can power exploitation. Audit your data practices and ask a blunt question: if our customers saw exactly how we use their data, would they thank us or feel betrayed? Ethical AI marketing means building the version you’d be proud to publish.
3. Transparency is a competitive advantage, not a compliance burden
DraftKings disputes the Times’ characterization. The company says its promotions are targeted based on continued use and customer engagement — not on losses — and points to standard metrics like customer lifetime value, retention, and return on investment.
Maybe. But here’s the problem: when the targeting logic is a black box, nobody can verify that claim — not customers, not regulators, and frankly not the public. And in the court of public opinion, “trust us, it’s engagement-based” is a losing argument once more than 40 former employees have described the system to the country’s paper of record.
Contrast this with brands that proactively explain their personalization. “We recommend products based on your past purchases — here’s how to control it.” That kind of transparency doesn’t weaken marketing; it strengthens it, because it gives customers a reason to trust you with their data in the first place.
The lesson: ethical AI marketing starts with explainability — document how your AI marketing tools make decisions, and be ready to explain it in plain English. If you can’t explain why a customer saw a particular ad, you have a transparency problem — and in 2026, transparency problems become PR problems fast. Ethical AI marketing treats transparency as a feature, not a risk.
4. Regulators are circling AI marketing — get ahead of them
The Times investigation notes that the revelations raise significant legal risks, with federal and state lawmakers actively refining AI compliance standards. This isn’t happening in a vacuum. Across the US, regulators are scrutinizing algorithmic targeting, dark patterns, and AI-driven personalization with growing intensity.
For small and mid-sized businesses, there’s a dangerous misconception that regulation only targets giants like DraftKings or Meta. In practice, enforcement actions create precedents, and precedents trickle down.
The lesson: build your ethical AI marketing practices as if they’ll be audited tomorrow. Clear consent flows, honest ad creative, accessible opt-outs, and no targeting based on vulnerability signals. Compliance isn’t just legal protection — it’s brand insurance.
5. Short-term ROAS can destroy long-term brand equity
Let’s talk numbers, because the business case for aggressive targeting always starts with numbers. According to the reporting, DraftKings distributed roughly $400 million through AI-automated promotions in 2025 alone, and executives said at an Investor Day in March 2026 that these automated incentives increased sportsbook promotional margins by 13 percent. The company reported $6.05 billion in revenue for 2025, up 27 percent year over year.
On a spreadsheet, the model was a triumph. In the real world, it’s now a national scandal — with “DraftKings AI” surging on Google Trends, congressional attention likely, and a brand association with the word “predatory” that no promotional margin can buy off.
The lesson: every marketing tactic should pass the headline test — how would this look on the front page of the Times? Tactics that juice this quarter’s return on ad spend while mortgaging your reputation are bad math. Sustainable growth — the kind ethical AI marketing is built for — comes from customers who trust you enough to come back. That’s the philosophy behind everything we do as a digital marketing agency committed to ethical AI marketing — growth beyond the numbers, not at the expense of them.
6. Build ethics into your martech stack, not on top of it
One reason the DraftKings story resonates is that it reveals how easily ethics become an afterthought. The targeting model was built, tested on roughly 5,000 casino players, refined, and extended to sports betting. The protective model was started later — and killed. Ethics wasn’t part of the system; it was a separate project competing for resources. It lost.
If you’re a business owner building an ethical AI marketing program in 2026, don’t treat ethics as a feature to add later. Build it into your selection criteria now: audit the objective function, check the inputs for vulnerability signals, demand explainability from vendors, and set kill criteria in advance.
The lesson: ethical AI marketing isn’t a slogan; it’s an operating procedure. Write it down, assign ownership, and review it quarterly like any other part of your growth strategy.
7. Choose marketing partners who’d pass the headline test
Here’s the uncomfortable truth for business owners: you may never build an AI targeting model yourself, but your agency, your ad platform, and your martech vendors make these decisions on your behalf every day. When you hire someone to run your Meta ads or manage your SEO, you’re trusting their judgment about where the line sits.
So ask them. Ask your agency how they use AI in audience targeting. Ask what data signals they exclude. Ask how they’d respond if a tactic worked brilliantly but felt wrong.
At KKeyQik, we’ve built our entire approach around this principle: data-driven marketing with a conscience. We use AI and automation to make campaigns smarter — better audience insights, sharper creative testing, faster optimization — but never to exploit the people we’re trying to reach. Measurable results and ethical methods aren’t opposites — that’s the whole point of ethical
AI marketing. They’re the only combination that lasts.
The consumer trust dividend: what the data says about ethical brands
Skeptics will ask the obvious question: does ethical AI marketing actually pay, or is it just a feel-good tax on growth? The data says it pays — and the gap is widening.
Study after study shows that consumers actively reward brands they trust with their data and punish brands they don’t. When people believe a company respects their privacy, they share more data voluntarily, opt into personalization more often, and forgive the occasional misstep. When they feel surveilled or manipulated, they install ad blockers, abandon carts, and churn — silently. You never see the revenue you lost to distrust; it just never arrives.
This creates a compounding effect that most ROAS-obsessed marketers miss. Exploitative targeting borrows conversions from the future: it squeezes a click today by spending trust you’ll need tomorrow. Ethical AI marketing does the reverse — every transparent, respectful interaction is a deposit. Over a two- or three-year horizon, the trusted brand has lower acquisition costs (more word of mouth, more returning visitors), higher lifetime value, and a customer base that actually wants to hear from it.
There’s a defensive angle too. Privacy regulation is tightening in the US every year, and platform policies shift constantly — just ask any advertiser who built their whole funnel on a targeting feature the platform later removed. Businesses that already market ethically are insulated from those shocks. The DraftKings approach is fragile by design: it depends on regulators, platforms, and the press never looking too closely. Hope is not a strategy.
The lesson: treat consumer trust as a balance-sheet asset. Measure it, invest in it, and report on it like revenue — because over time, it becomes revenue.
5 AI ad targeting red flags to ban from your campaigns
If you want a practical starting point, here are five tactics to prohibit explicitly — in your own team and in any agency agreement:
- Targeting distress or vulnerability signals. Browsing patterns that suggest financial stress, health anxiety, addiction, or desperation should be exclusion criteria, never targeting criteria. If a segment converts “suspiciously well,” ask why before you scale it.
- Lookalikes built on your most compulsive buyers. “Find me more customers like my top 1%” sounds smart until you realize your top 1% includes people who can’t stop buying. Seed lookalike audiences from your healthiest customers — high satisfaction, low refund rates — not just your highest spenders.
- Dark patterns in consent and opt-outs. Pre-ticked boxes, maze-like unsubscribe flows, and “accept” buttons ten times bigger than “decline” are the consent equivalent of fine print. Regulators are specifically targeting these in 2026.
- Frequency without mercy. AI-driven retargeting can follow a user across the entire internet for weeks. Set hard frequency caps and cooling-off periods — especially for high-consideration or sensitive purchases. Stalking is not a funnel strategy.
- Black-box “advantage” features with no explainability. If neither you nor your agency can explain why the algorithm showed an ad to a specific person, you can’t defend it either. Demand transparency from every AI ad targeting tool you pay for, or replace it.
Ban these five and you’re already marketing more ethically than most of your competitors — and sleeping better than DraftKings’ PR team.
How to audit your own ethical AI marketing: a 5-minute checklist
You don’t need a data science team to apply these ethical AI marketing lessons. Run through this checklist this week:
1. List every AI-powered marketing tool you use — ad platforms, email personalization, chatbots, recommendation engines.
2. For each one, write down what it optimizes for in one sentence. If you can’t, that’s a red flag.
3. Identify the data signals feeding each tool. Flag any that could indicate vulnerability.
4. Check your consent and opt-out flows. Can a customer easily understand and control how they’re targeted?
5. Run the headline test on your most aggressive campaign. If it would embarrass you on the front page, change it before someone else writes that story.
The bottom line
The DraftKings story isn’t really about gambling. It’s about what happens when powerful technology meets unchecked incentives — and it’s a preview of the reckoning coming for AI marketing across every industry.
The businesses that thrive in the next decade won’t be the ones with the most aggressive targeting. They’ll be the ones practicing ethical AI marketing — the ones customers trust with their data, their attention, and their wallets. That trust is built campaign by campaign, decision by decision — starting now.
If you want a marketing partner that grows your business without crossing lines — a team built on ethical AI marketing — talk to our team. We’d rather build you a brand that lasts than a spike that scandals.
Frequently Asked Questions
A 2026 investigation reported that DraftKings built a machine-learning system ranking casino players by expected losses, then targeted the highest-loss players with promotional incentives. The story is a cautionary tale about AI optimization without ethical boundaries.
Ethical AI marketing means constraining what your systems optimize for — not just celebrating how well they optimize. AI has objectives, not values, so businesses must set explicit boundaries around vulnerability signals and exploitative targeting.
The same targeting technology is available to every advertiser on major ad platforms. The question is no longer whether you use AI in marketing, but how you use it and where you draw the line.
Treat every campaign like a small policy decision: write down the customer experience it creates, who it excludes, and what could look uncomfortable if disclosed. If a tactic cannot survive that short written review, scale it back or drop it. This habit usually costs an hour per campaign and prevents trust damage that takes quarters to repair.
Platforms keep adding new AI targeting features, and a feature you never asked for can quietly change how your ads are served. A short review every quarter catches those shifts before they run for months. Put it on the calendar alongside your other marketing reviews so it never depends on memory.
Ask for the opt-out path first: how customers can avoid it, how you can turn it off, and who owns the decision. Features with no clear off switch tend to expand quietly across campaigns. Writing that down before launch keeps your team in control of the scope.
A one-page note per campaign: the tool used, the objective, the data signals, and how a customer can opt out. Keep it in a shared folder your team can hand to a client or partner on request. When questions arise later, that record turns a scramble into a five-minute lookup.
Three basics cover most questions: your consent and opt-out flow records, a list of the AI tools you use with what each optimizes for, and your vendor agreements on data use. Keep them in one place and review them when platforms change features. Most compliance friction comes from hunting for records, not from the rules themselves.
Spell out the non-negotiables: excluded audience signals, frequency caps, consent requirements, and your right to pause any campaign that crosses a line. Ask for plain-language reporting on what the AI optimized for each month. Contracts turn good intentions into enforceable standards.
Watch opt-in rates for personalization, repeat purchase frequency, review sentiment, and the volume of complaints about your ads or targeting. When those move in the right direction together, trust is doing its job as a growth lever. These signals are slower than click-through rates, but they predict the revenue that lasts.