A New York Times investigation exposed something genuinely damning for the online gambling industry. DraftKings built an AI model to identify which customers would lose the most money, then spent roughly $400 million steering promotions toward exactly those people.
On its own the story has serious ramifications, but the more important question it raises isn’t about DraftKings specifically, it’s about a much larger part of the gambling industry where a story like this would almost certainly never come to light.
The DraftKings AI Model
DraftKings built a machine-learning model in 2023, internally called an “elasticity score”. It predicted how much a customer would lose after receiving a bonus or free bet, based on play frequency, account balance patterns, and loss-to-wager ratios.
What the Model Was Used For
| Detail | Reported Finding |
| Model | “Elasticity score” |
| Built | 2023 |
| Purpose | Predict customer losses after bonuses or free bets |
| 2025 promotions | Roughly $400 million |
| Target | Accounts projected to keep losing the most |
In 2025 alone, the company directed roughly $400 million in promotions toward the highest-scoring accounts: the customers projected to keep losing the most.
A former data analyst who worked on the model told the Times directly:
“The best investment would be a problem gambler.”
A separate model, built specifically to flag gamblers heading toward crisis, was reportedly shelved rather than deployed.
DraftKings disputes the framing, saying its promotions target engaged, active users rather than people selected by expected losses.
How Was DraftKings Caught
DraftKings is a publicly traded US company, filing disclosures with the SEC, subject to state gambling regulation in dozens of jurisdictions, employing thousands of people who can leave and talk to reporters.
The Times built this story on internal memos, betting records, and more than 40 former employees willing to go on record.
Why That Investigation Was Possible
Public company
2SEC disclosures
3State gambling regulation
4Thousands of employees
5Internal records + former employees
6Investigative reporting
None of that exists by accident.
It exists because DraftKings operates inside a system with enough transparency, enough regulatory exposure, and enough of a public paper trail that a serious investigation was actually possible.
Now consider what that same investigation would require at a typical offshore crypto casino.
Beneficial ownership is frequently undisclosed or hidden behind nominee directors.
There’s no SEC filing, no public earnings call and often no fixed jurisdiction of meaningful legal consequence.
| DraftKings | Typical Offshore Crypto Casino |
| Publicly traded | Often privately held |
| SEC filings | No equivalent SEC disclosure |
| State gambling regulation | Offshore licensing |
| Identifiable workforce | Ownership and workforce may be difficult to identify |
| Public paper trail | Often limited public disclosure |
| Former employees available to reporters | Potentially difficult to identify or locate |
Curaçao and Anjouan licensing, the most common regulatory homes for crypto casinos, don’t carry anything resembling the compliance infrastructure, public disclosure requirements, or enforcement capacity of a state gambling commission.
There’s no equivalent of 40 former employees, because there’s often no meaningfully identifiable company to have been employed by in the first place.
The DraftKings headquarters in Boston’s Back Bay neighborhood. The company’s physical corporate presence provides an identifiable point of accountability that is often harder to establish with offshore operators.
This is a structural point, not an accusation
There’s no confirmed evidence that any specific offshore crypto casino runs a comparable loss-maximization model.
We are not making any such claim here. The actual point is different and, in some ways, more unsettling.
The reason DraftKings’ practice became known isn’t that DraftKings is uniquely bad among gambling operators, it’s that DraftKings operates inside one of the only parts of this industry where a practice like this could become known at all.
Absence of a similar investigation into offshore crypto casinos isn’t evidence that nothing similar happens there.
It’s evidence that the infrastructure required to find out such as disclosure obligations, identifiable employees, regulatory audit trails, largely doesn’t exist in that part of the industry to begin with.
AI itself isn’t necessarily the problem
AI applied to gambling isn’t inherently harmful. The same underlying technology can genuinely improve a player’s experience.
AI-driven game recommendations, adaptive interfaces, and predictive features that make navigating a casino platform smoother are already in real use at several crypto casinos.
AI is also being used to support responsible gambling by gently surfacing tools like session timers or deposit limits when they might actually help a player.
The same technology can:
Flag a player sliding toward harm
Offer a deposit limit
Surface responsible-gambling tools
or
Identify players likely to lose the most
Target those players with bonuses
Encourage further play
This reflects the double edge sword of an advanced technology such as AI.
The same type of tool that could flag a player sliding toward harm and offer them a limit was used by DraftKings to finds those same players and send them more bonuses.
The technology was never the obstacle. The choice was.
How this connects to other gaps in the industry
This isn’t a new theme.
CFTC-regulated prediction markets like Kalshi and Polymarket fall entirely outside responsible-gambling requirements, a direct consequence of their legal classification rather than their actual risk to users.
Self-exclusion tools like GamStop or Spelpaus are comprehensive within licensed markets and simply don’t extend to offshore platforms at all.
| Example | Gap |
| CFTC-regulated prediction markets | Outside responsible-gambling requirements |
| GamStop / Spelpaus | Limited to licensed markets |
| Offshore platforms | Limited accountability and disclosure |
Each of these is a different specific mechanism, but they all point at the same underlying fact.
Regulation, imperfect as it clearly is even at a company as scrutinized as DraftKings, still produces real accountability that offshore, anonymously-owned platforms structurally don’t have to answer to at all.
What this actually means for players
If a fully regulated, publicly traded operator that files disclosures, answers to state regulators, and employs thousands of people could build a system that specifically chased its biggest losers—and only get caught after years, through an investigation that required dozens of former employees to talk—then the honest baseline for an anonymously owned offshore platform with no equivalent disclosure obligations should not be more trust.
It should be less. Not because every offshore operator is doing the same thing, but because there is often no comparable mechanism by which you would ever find out either way.
That is not a reason to assume the worst about any specific casino.It is a reason to weigh the things you can verify such as:
License
Ownership transparency
Real dispute-resolution options
And it is a reminder that every casino, regulated or not, is in the business of making money. Protect yourself accordingly by using deposit limits, session timers, and self-exclusion tools wherever they are available.
The post DraftKings Got Caught. Offshore Crypto Casinos Almost Never Will. appeared first on BitcoinChaser.

1 month ago
13

Bengali (Bangladesh) ·
English (United States) ·