6th Week

• Tech firms’ value comprises a quarter of the US stock market value

• Top 5 more than half of this: Microsoft, Amazon, Apple, Alphabet, Facebook

• This has increased further as Covid forced more activity online

• 20 ears old

• Grown to a behemoth of more than $150bn revenues and about $1trn market cap

• Google’s share of the US ad market reached 30%

• EU antitrust cases did not leave much of a dent: The first (self-preferencing in search) barely affected market capitalization, the second (exclusivity of Android and default Google Search) and third (exclusivity for AdSense) did not affect its rise

• Google launched Android (2008) to control the mobile ecosystem — search, ads, and data.

• Strategy: open system inviting many hardware makers and app developers.

• Contrasts with Apple’s closed “walled garden” model.

• Result: Android dominated market share, overtaking Symbian and BlackBerry.

What are the motives?

• Complementary innovation?

• Complementary businesses?

• Elimination of potential competitors?

• Google announced its Fitbit acquisition in 2019 to expand into health data.

• The EU investigated (2020) over concerns it would strengthen Google’s dominance in online advertising by combining health and fitness data.

• Concerns focused on data use and interoperability with other devices, not hardware.

• The merger was approved in Dec 2020, with commitments on data access and privacy safeguards.

• In 2020, the US FTC asked major tech firms (like Google) for data on all acquisitions since 2010.

• Many deals were below notification thresholds, escaping review.

• The FTC aimed to assess whether big techs buy potential competitors (“killer acquisitions”).

• Purpose: determine if smaller acquisitions should face premerger reporting requirements.

• Facebook bought Instagram (2012) and WhatsApp (2018, $18B) to secure dominance in social media and advertising.

• These platforms were potential competitors for user attention and ad space.

• Acquisitions helped Facebook retain younger audiences as they moved away from its main app.

• Antitrust regulators didn’t act at the time—raising questions about oversight and competition.

• FTC launched an anti-monopolization case in 2020, still ongoing.

Few firms are getting bigger and bigger:

• Zero MC -> return to scale

• adoption of online news, shopping social Networks

• direct and indirect network effects

• growth of large digital ecosystems

New:

• Global reach + global tax evasion

• Lots of acquisitions of smaller or not so small firms

Treaty on the Functioning of the European Union has specific provisions on competition policy

• TFEU Article 101: anti-competitive agreements

• TFEU Article 102: abuse of dominant position

The European Union also intervenes in merger control, in order to avoid that

mergers harm competition and consumers

Prohibits:

• Agreements that prevent, restrict, or distort competition in the EU market.

• Examples include:

  • Fixing prices or trading conditions

  • Limiting production, markets, or innovation

  • Sharing markets or supply sources

  • Applying unequal terms to equivalent deals

  • Imposing irrelevant contract obligations

Exemptions:

• Allowed if agreements improve production, distribution, or innovation

• Must benefit consumers, avoid unnecessary restrictions, and not eliminate competition

Prohibits:

• Any abuse of a dominant market position that affects trade between EU Member States.

Examples of abuse:

• Unfair pricing or trading conditions (exploitative behavior)

• Limiting production or innovation to harm consumers

• Discriminating between equivalent partners (unequal treatment)

• Tying or imposing irrelevant obligations in contracts

Goal:

Prevent powerful firms from using dominance to distort competition and harm consumers.

Aim:

Prevent mergers that reduce competition in the EU market.

Key Rules:

• Mergers above a certain size (based on revenue/market share) must be reported:

  • To national competition authorities, or

  • To the European Commission (DG Comp) for large cross-border cases.

Investigation Phases:

1. Phase 1: Check for potential competition problems — approve if none.

2. Phase 2: Detailed review — may block, or approve with conditions (commitments by firms).

Position of strength in a market that allows a firm to behave to an appreciable extent independently of its clients and competitors

• not illegal as such

• But abuses of dominance are forbidden, such as

  • Foreclosure or predation of competitors, excessive pricing or imposition of unfair trading conditions, leverage of dominance into other markets

• Google in search advertising

• Complaint by DOJ 2020 about exclusion of rivals

• Booking.com in hotel bookings

• Several court cases about “most-favourite-customer” clauses

• Apple Store and other app stores:

• Accusation of excessive fees of up to 30%

Asimov’s Three Laws of Robotics (1942):

• A robot may not harm a human or allow harm through inaction.

  1. A robot must obey human orders unless they conflict with the First Law.

  2. A robot must protect its own existence unless it conflicts with the first two laws.

     
     

• Originally about physical robots, now relevant to algorithms and AI in the “cloud.”

• Today, algorithms influence society and markets → need for consumer protection and competition policy in AI governance.

• AI/Algorithmic recommendations heavily influence what we watch on Youtube, buy on Amazon, see on Facebook

• Most people feel powerless to control fake news and rarely try to stop it.

• Fake news creates widespread confusion.

• Social media platforms, the public, and politicians all share blame for not addressing it.

• Algorithms amplify — they are designed for spreading content, not stopping misinformation.

• Uses machine learning to analyze individuals’ preferences and willingness to pay.

• Benefits: personalized offers, better product matches.

• Harms: higher prices, loss of consumer surplus.

• Two types (Bar-Gill, 2019):

  
  

  • Preference-based PD: hurts consumers but increases efficiency.

  • Misperception-based PD: both harms consumers and reduces efficiency.

  
  

• Proposed fix: legal tools or personalized law to counteract unfair pricing.

• Cartels = explicit agreements between firms to limit competition and innovation.

• Purpose: create market power, raise prices, increase profits.

• Illegal — they block market functioning and harm consumers.

• Competition authorities actively hunt and penalize cartels.

————————

• Tacit collusion = unspoken coordination — firms don’t compete because they’ve learned that aggressive behavior triggers punishment (e.g., price wars).

• Not illegal, since there’s no explicit agreement, but outcomes are similarly harmful (higher prices, less innovation).

• Regulators try to prevent it indirectly, e.g. by avoiding mergers that make collusion easier (fewer firms = higher risk).

• Algorithms can learn to coordinate prices (avoid price wars, keep prices high) without explicit communication or programming.

• This may lead to tacit collusion facilitated by AI — a new challenge for competition authorities.

• Legal dilemma: if no human intent or agreement exists, who is guilty of collusion?

• Voluntary codes of conduct → Respect privacy, limit data sales.

• Transparency rules → Algorithmic codes reviewed by competition authorities.

• Obligations/prohibitions → e.g. delete personal data, tackle fake news.

• Liability laws debated — focus on illegal content and anti-competitive outcomes, not just conduct.

European Union: creation of an interlocking set of Regulations

• Ex-ante rules instead of ex-post competition enforcement: DMA, DSA

• Frameworks for data spaces: Data Act and Data Governance Act

• Ex-ante rules for Artificial Intelligence: AI Act

After many years of discussions, the Digital Markets, Competition and Consumers (“DMCC”) Act became law in May 2024

• Issues: Google/Facebook dominance → higher prices, media decline, weak data control.

• Fix: Create Digital Markets Unit (DMU) with ex-ante powers and a Code of Conduct for Strategic Market Status firms (fairness, openness, transparency).

• Tools: Data access, interoperability, transparency — and platform break-up as last resort.

• Law: DMCC Act (2024) gives CMA power to enforce firm-specific remedies.

The single most important competition policy initiatives of the 2019-2024 European Commission, setting rules for the digital economy

• Aim: Provide tools to the European Union to enjoy the benefits of the digital economy, while protecting consumers and stopping abuses of market power

• Traditional ex-post competition enforcement much too slow! (Art. 101/102 TFEU)

• DSA: rspns & transparent digi markets -> Focus on relation between users and platforms

• DMA: reign in the power of gatekeepers -> Focus on relationship between platforms, competitors and business clients

• Applies from: January 2024

• Scope: All online intermediaries (lighter rules for SMEs/startups)

• Structure: Hierarchy of services — intermediation → hosting → platforms → Very Large Online Platforms (VLOPs) and Very Large Online Search Engines (VLOSEs)

• Examples:

  • VLOPs: Amazon, AliExpress, YouTube, Instagram, Wikipedia, Shein, Temu, etc.

  • VLOSEs: Google Search, Bing

  
  

1. Safe & Accountable Online Environment – Prevents illegal/harmful content and disinformation; ensures user safety and rights.

2. Protection of Consumers & Fundamental Rights – Clear rules for online rights, stronger consumer and child protection.

3. Fostering Innovation & Competitiveness – Supports SMEs/startups; promotes innovation and fair growth.

4. Rebalancing Roles – Clarifies responsibilities of users, platforms, and authorities; citizens at the centre.

5. Legal Certainty & Single Ruleset – Harmonized EU framework for easier business operation.

6. Access to EU-wide Markets – Ensures fair access and legal harmonization across member states.

7. Democratic Control & Oversight – Strengthens oversight over large platforms; tackles manipulation and disinformation.

• Applies since: May 2023

• Goal: Prevent abuse of power by “gatekeepers” (big tech) against competitors and business clients → ensure fairness & contestability.

• Gatekeepers: Alphabet, Amazon, Apple, ByteDance, Meta, Microsoft (26 core platform services like search, ads, app stores, social media, OSs, cloud).

• Rules: Must comply with DMA obligations by March 2024.

• Definition: Based on EU users, turnover, market cap; includes a process to contest gatekeeper status.

Obligations:

• Notify EU when acquiring firms (merger control)

• Allow users to uninstall gatekeeper apps

• Ensure data portability and access for clients and users

Prohibited practices (ex-ante):

• Self-preferencing in search/intermediation

• Using marketplace data against sellers

• Exclusive pre-installation or marketplace exclusivity

• Combining user data across core services

Sanctions:

• Fines, remedies, or break-up of EU operations

Greylist (to be investigated):

• Blocking third-party sellers from customer data

• Collecting unnecessary personal data

• Restricting hardware makers from offering alternative apps/services

Ongoing enforcement:

• EU Commission monitors and takes monthly action under the DMA

• Applies from: Sept 2023

• Goal: Enable voluntary data sharing (industrial & IoT) → reduce “data hoarding”

• Scope: Both personal & non-personal data → creation of European Data Spaces

• Introduces:

  • Rules for data intermediation services

  • Re-use of publicly held protected data (under Open Data Directive)

  • Data altruism organisations & European Data Innovation Board

• Covers: Access, payments, and international data flows

• Goal: Framework for data access and use → define who can use what data and under which conditions

• Key features:

  
  

  • Fair value allocation & innovation incentives

  • IoT users gain control over generated data

  • Data sharing duties between businesses

  • Interoperability rules for data flows across EU sectors

  • Fair cloud competition & protection from unfair contracts

  • Public sector access allowed in emergencies only

  • Safeguards against non-EU government access to EU data

  
  

• In force: August 2024

• Bans: High-risk AI uses (e.g. social credit scoring, biometric surveillance)

• Sets: Minimum safety and transparency rules for AI in sensitive sectors

• Regulates: General-purpose AI models (e.g. OpenAI, Anthropic)

• Uses: Risk-based classification via compute thresholds

• Governance: EU AI Office + national regulators oversee compliance

Risk classification of AI systems:

• 🟥 Unacceptable risk: Prohibited — e.g. biometric surveillance, social scoring

• 🟧 High risk: Strict rules — risk assessment, mitigation, and oversight for AI impacting jobs, credit, border control, justice, infrastructure, etc.

• 🟦 Limited risk: Transparency required — users must know when interacting with AI (e.g. chatbots, image/text generators)

• 🟩 Minimal risk: No obligations — e.g. spam filters, video game AIs