It's nuts that this is all they can manage. Like, take 30% or 40% -- that might actually encourage compliance.
I think they'll comply, but almost entirely not because of small fines.
It's nuts that this is all they can manage. Like, take 30% or 40% -- that might actually encourage compliance.
I think they'll comply, but almost entirely not because of small fines.
Like if you're building gambling systems for 7 year olds then yes I think you should lose your job and then some. You the programmer, your product manager, your people manager, all the way up the chain. Everyone who participated should be individually accountable.
I'd strongly argue that 6% will not correct bad behavior.
What corrects bad behavior? The threat of serious consequences to a business.
For a company like Google this fine would be in the vicinity of $24 billion, which would be a bit over 20% of their annual profit.
So for companies that have to answer to shareholders this is already a pretty big deterrent, but just in case if they reject compliance it becomes a periodic penalty!
We've seen Zuckerburg pivot from openly pledging to defy the EU [1] and then presenting compliant plans six months later to avoid the the periodic fine [2].
But we've also seen Apple in a bid to avoid the periodic fine imposing a ~20% commission [3] on third party payments after their 27% fee was ruled illegal and they were told it must be free [4]. The investigation into that has taken a year now despite being immediately perceived as noncompliant by many developers and ressembling what Google was recently fined €430 million for [5] and Apple seems to feel no pressure to change it.
I guess what happens next to Apple will decide if companies fear this regulation, the periodic fine they were due if they didn't drop the 27% fee was touted to be up to $50m/day retroactively for the entire noncompliance period, which is now at least 400 days so lots of opportunity to send a message.
[1] https://www.reuters.com/sustainability/boards-policy-regulat...
[2] https://www.reuters.com/business/meta-offer-choices-personal...
[3] https://daringfireball.net/2025/06/apple_app_store_policy_up...
[4] https://ec.europa.eu/competition/digital_markets_act/cases/2...
[5] https://www.theguardian.com/technology/2026/jul/23/eu-fines-...
If your company is significantly breaking the law, should the company have profit for that year?
Paying the bills is one thing, but profit to shareholders?
Not an accurate analogy. This is a fine on sales, not expenditures. It would be like fining you 6% of your pre-tax income, not doubling sales tax on things you buy.
> I'd strongly argue that 6% will not correct bad behavior.
If the government threatened to take 6% of your pre-tax income if you violated a set of laws, would you laugh it off and refuse to change your behavior? Or would you make an effort to not incur those fines?
If that was true, sales would fall to 67% of the pre-law sales. That doesn’t happen. It’s a silly claim.
It would be like the government taking 6% of your total compensation for breaking the speed limit. You could try to speed everywhere to save time but if you get caught it’s going to hurt.
It’s not a “increase sales 50%” scenario. You’re arguing against something that doesn’t exist.
6% of annual, global sales, as an extra cost from one jurisdiction, which can impose this multiple times if there are multiple breaches, and where they're not the only jurisdiction with such rules, and the DSA isn't the only ruleset where fines are based on global revenue even in the EU (famously, GDPR).
Or, to put it another way: 16 such fines in one year and you've lost 100% of your revenue. You don't get to make up for this by raising prices, because that just means "100% of your revenue" is now a bigger number.
Sometimes it’s more profitable to eat the fine than change your behaviour.
Than from the total tax income from Europe's own public tech companies!"
frist_fine = 0.06 × anual_global_sales
fine(n) = first_fine * n^x
And the exponent x is to be chosen by how harsh you want to be. Repeat offenders will quickly run out of money as long as your exponent is reasonably bigger than 1.If we take an exponent of 2 a repeat offense by OpenAI would look like this:
1st offence: EUR 1.50 billion (6%)
2nd offence: EUR 6.00 billion (24%)
3rd offence: EUR 13.50 billion (54%)
4th offence: EUR 24.00 billion (96%)
5th offence: EUR 37.50 billion (150%)
6th offence: EUR 54.00 billion (216%)It is hard to "budget in" a expense that is multiple time your earnings. Especially if that expense is growing exponentially each time..
But shows that people still don't understand exponential functions, even on this platform. We are truly doomed.
"We lose money on every deal, but we plan to make it up in volume" was one of the infamous quotes from the dot.com bubble.