Justiz Bayern, hier.
Friday, July 31, 2026
Thursday, July 30, 2026
Tuesday, July 28, 2026
Sunday, July 26, 2026
Epic v. Google is settled: what's next
"The parties are *advised* that the Court will retain Professor Nancy L. Rose to assist the Court in monitoring Google’s compliance with the injunction'
She was instrumental in preventing a gigantic "Epic Fail" - ICYMI, but nothing can replace public enforcement in the public interest.
Saturday, July 25, 2026
Modern Economic Analysis and Antitrust Law: A Guide
S. Salop, here.
From what I have seen, this "modern" economic analysis certainly moves beyond a narrow focus on price. It takes innovation, quality, variety, entry and dynamic effects much more seriously, and it is far more sceptical than the Chicago tradition of claims that monopoly necessarily promotes innovation.
Yet it still appears to remain fundamentally efficiency-based. Competition, diversity and market openness matter chiefly because they are expected to generate better welfare outcomes. This means that exclusion may still be justified if the dominant firm can show sufficiently large efficiencies and persuade the decision-maker that those efficiencies outweigh the loss of rivalry.
That is precisely where a much deeper change of paradigm is needed. The issue should not merely be whether a single gatekeeper can organise the market more efficiently than a plurality of competing actors. It should also be whether one private firm ought to possess the power to determine the entire structure of market access, the permissible forms of innovation and the conditions under which others may compete.
Contestability, diversity and independent routes to market should therefore not be treated simply as variables in an efficiency balance. They are constitutive elements of a competitive order. The objective should not be merely to maximise welfare within a market structure controlled by the incumbent, but to preserve the possibility of decentralised experimentation, rival forms of organisation and challenges to the incumbent’s model.
In that sense, this type of newish economic analysis is not yet the necessary paradigm shift. It offers a more sophisticated balancing exercise when what may be required is a structural presumption against private control over the architecture of competition itself.
Yet it still appears to remain fundamentally efficiency-based. Competition, diversity and market openness matter chiefly because they are expected to generate better welfare outcomes. This means that exclusion may still be justified if the dominant firm can show sufficiently large efficiencies and persuade the decision-maker that those efficiencies outweigh the loss of rivalry.
That is precisely where a much deeper change of paradigm is needed. The issue should not merely be whether a single gatekeeper can organise the market more efficiently than a plurality of competing actors. It should also be whether one private firm ought to possess the power to determine the entire structure of market access, the permissible forms of innovation and the conditions under which others may compete.
Contestability, diversity and independent routes to market should therefore not be treated simply as variables in an efficiency balance. They are constitutive elements of a competitive order. The objective should not be merely to maximise welfare within a market structure controlled by the incumbent, but to preserve the possibility of decentralised experimentation, rival forms of organisation and challenges to the incumbent’s model.
In that sense, this type of newish economic analysis is not yet the necessary paradigm shift. It offers a more sophisticated balancing exercise when what may be required is a structural presumption against private control over the architecture of competition itself.
The central distinction is between better balancing within the welfare paradigm and treating plurality, contestability and dispersed power as values that cannot be traded away for claimed efficiencies.
What is needed is a true modern (new) competition school, not just an incremental fixing of the same, inadequate economic paradigm.
Thursday, July 23, 2026
Commission fines Google €890 million for breaches of the Digital Markets Act
EC, here.
"The Commission also takes note of Google's proposals on how it plans to apply the principles of the decision to AI Overviews and AI Mode, on which dialogue will continue in light of today's decision"
This is broadly the direction anticipated here (no crystal ball, just common sense). The Commission has now placed AI Overviews and AI Mode explicitly within the Article 6(5) dialogue. It should move quickly. Google is already deploying these services through Search, and the underlying distribution issue is sufficiently clear. A specification proceeding would provide a timely opportunity to define the conditions governing their integration, prominence and access to users. The dialogue on this topic is already well underway, as the press release notes; allowing it to remain informal would merely give the architecture more time to harden.
Wednesday, July 22, 2026
UK's Bloomsbury among beneficiaries of $1.5 billion Anthropic copyright lawsuit settlement
Reuters, here.
Too little, too late?
Google's Fuite en Avant? A DMA Response Becomes Harder to Avoid
France had so far remained outside the rollout, apparently because of the unresolved dispute over press content and publisher choice. In 2024, the Autorité de la concurrence fined Google €250 million for breaches of commitments governing negotiations with publishers. Google now refers to constructive discussions with authorities and publishers. The new framework reportedly includes a separate opt-out, specific information on impressions generated by the AI services, and remuneration for publishers whose content is used.
It remains unclear whether the framework fully meets the Autorité de la concurrence’s expectations, and further adjustments may follow. This looks less like a small victory than a unilateral attempt by Google to create facts on the ground in France. Google appears to have concluded that its revised arrangements are sufficient under the existing commitments, leaving the Autorité to assess them after deployment. Most importantly, this leaves Google’s chosen architecture intact, including the use of Google Search as the distribution infrastructure for its own AI services.
A more impactful, low-hanging fruit is already on the Commission’s table: Article 6(5) DMA. The Commission has a pending non-compliance proceeding concerning preferential treatment within Google Search and issued preliminary findings in March 2025. That proceeding currently concerns Google’s own vertical services. On that (too long) existing front, we may soon see the non-compliance decision, probably accompanied by a rather large fine. AI Overviews and AI Mode raise the same distribution issue on a broader scale: whether Alphabet may use Google Search to give its own answer services privileged prominence and immediate access to users. A dedicated specification proceeding might, indeed, be particularly useful here: it could define the conditions under which Alphabet may integrate and promote AI Overviews and AI Mode within Google Search. The Commission has just completed two specification proceedings concerning Google, on Android interoperability and access to Search data. Whether it is prepared to open a third will indicate how seriously it intends to treat AI integration as a contestability problem.
Tuesday, July 21, 2026
Saturday, July 18, 2026
Friday, July 17, 2026
Spaghetti Revolving Doors à la Meloni
The leadership of Italy’s Competition Authority has just been entrusted to a figure whose family background and institutional trajectory are deeply rooted in the Italian extreme right. Former Clearly Gottlieb, who has been representing Google & Co.
Now we watch.
Thursday, July 16, 2026
Like the Red Queen, the DMA in the AI age must run as fast as it can merely to remain in the same place.
Commission provides guidance to Google for AI interoperability on Android and sharing of Google Search data under the Digital Markets Act - EC, here.
Much tbd. Decisions not yet available, just a few thoughts for now.
The Article 6(7) specification proceeding may represent a transitional stage in DMA enforcement. The functional capabilities identified by the Commission—invocation, contextual awareness, the performance of actions and access to system resources—increasingly characterise the AI-assistant layer rather than the operating system as such. As AI assistants become the primary interface through which users access digital ecosystems, enforcement may therefore need to move from regulating operating-system interoperability to regulating AI assistants as core platform services in their own right.
Such a shift may already be possible under the existing DMA. Virtual assistants are expressly included among the categories of core platform services, and that concept could be interpreted functionally and in technologically neutral terms. This would be consistent with the General Court’s recent judgment in Apple, which rejected an understanding of a core platform service determined by the particular device or operating system through which it is provided, focusing instead on the common function performed by the service.
Should the Commission consider that interpretation legally uncertain, however, it should pursue the safer legislative route: initiating the procedures required to recognise generative and agentic AI assistants explicitly as a distinct category of core platform service. The DMA itself provides for investigations into new digital services and for periodic reconsideration of the existing CPS list, including through legislative proposals. Digital regulation cannot preserve its relevance by standing still. To borrow from the Red Queen, it must do all the running it can merely to remain in the same place; otherwise, the DMA risks becoming not the framework governing the emerging digital ecosystem, but an archaeological layer buried beneath it.
Wednesday, July 15, 2026
Rapport de la commission d’enquête sur nos vulnérabilités et dépendances numériques
Assemblée nationale, ici.
Monday, July 13, 2026
Saturday, July 11, 2026
What kind of incentives are we creating for the future of journalism, entertainment, and original content
C. Pedigo, here.
At Politico Tech Week I argued this isnthe type of questions useful economic thinking should deal with - forget Industrial Organization, which is mostly only a dangerous distraction.
Dreading your next teaching semester in the AI age? AI-resilient pedagogy might help
U. Chicago, Law School, here.
I subscribe to many of the points made here.
However, devices are great to take notes. Condemning students to the use of pen and paper is anachronistic at best. I'd erect barricades to restore the right to use my pad and e-pen in the classroom!
For IO economists and consultants only? What could possibly go wrong?
Theories of dynamic harm
- Giulio Federico, Oxera, Project Director for the Study
- Chiara Fumagalli, Bocconi University
- Nancy Rose, MIT
- Michele Polo, Bocconi University
- Laurent Eymard, BRG
Wednesday, July 08, 2026
A gatekeeper cannot transform a challenge to designation into a general anticipatory attack on the whole DMA regulatory regime.
Yours, EU GC here.
Also extremely interesting: "Since Apple has been unsuccessful, it must be ordered to pay the costs incurred by the Commission and Coalition for App Fairness, in accordance with the form of order sought by them."
Top: "the EU legislature intended to prevent the classification of a service under the DMA from varying according to the technological choices made by the gatekeeper" - technology neutrality. E.g., relevant legal question: does an AI assistant perform the function of a virtual assistant under the DMA?
Re EECC: hat the Court anticipated possible Apple's moves like a well-chosen national proceeding (before the BNetzA?) that generates a preliminary reference on Article 2(7) - eventually produce a CJEU ruling binding across both regimes?
Tuesday, July 07, 2026
Monday, July 06, 2026
Thursday, July 02, 2026
Google Android: the Court of Justice upholds Google’s fine of around €4.1 billion
CJEU, here.
Judging from the press
release, this is a victory on every front. And, this time, quite
literally without appeal. Whether EU competition law can still be
rescued and made more effective is another matter, but it should at
least put some courage back into our enforcers and policy makers!
Favourite
sentence so far "This applies all the more where the conduct at issue
does not lend itself to a quantitative, price-based analysis".
Least favourite: BEUC - The European Consumer Organisation should
bear its own legal costs. If the EU is serious about civil society as
part of its enforcement infrastructure the economics of standing before
its courts deserve the same scrutiny as the substance argued there. The
courts anchors liability in consumer harm (e.g., para. 414 on diversity
of information sources) but their representatives see nothing of the
€4.125 billion fine and have to bear their legal expenses?
Interestingly,
the Stockholm/PriceRunner ruling invites Android claimants (rival
search services, browser vendors, OEMs even?) to plead continuation past
2018. This could at least provide solid economic incentives to comply
with the related DMA obligations - quickly.
Wednesday, July 01, 2026
Subscribe to:
Posts (Atom)