Humanein the Loop
← The matrix
07

AI power should be balanced in society

Current path

AI power concentrates in a handful of firms with unprecedented influence over markets, public discourse, and policy. Open alternatives are starved of capital and compute. Democratic institutions struggle to hold AI firms accountable.

Better future

AI power is distributed across a healthy ecosystem of commercial, open-source, public, and cooperative actors. Antitrust and competition policy apply. Democratic institutions retain capacity to govern.

Drift across the three domains

Norms

Regressing43 signals
CHT recommends
  • Treat AI market concentration as a competition and democratic concern.
  • Sustain substantive public debate on open vs closed AI beyond either-side capture.
Indicators we track
  • 7.N.aAntitrust and competition discourse applied to AI
  • 7.N.bOpen-source vs closed-source debate
  • 7.N.cConcerns about democratic capture

Laws

Advancing33 signals
CHT recommends
  • Apply antitrust and competition rules to AI markets.
  • Mandate interoperability and data portability.
  • Fund public-option AI and sovereign compute alternatives.
  • Restrict political use of AI (deepfakes, AI-generated campaign speech).
Indicators we track
  • 7.L.aAntitrust action against AI market concentration
  • 7.L.bInteroperability and data portability mandates
  • 7.L.cPublic-option AI and sovereign compute funding
  • 7.L.dRestrictions on political uses of AI

Design

Advancing31 signals
CHT recommends
  • Support responsible open-weights releases of capable models.
  • Invest in decentralized and federated architectures.
  • Provide fair API access to closed models for third-party developers.
Indicators we track
  • 7.D.aOpen-source model releases
  • 7.D.bDecentralized and federated architectures
  • 7.D.cThird-party access to closed models

Recent signals

RegressingMajorDesign · 7.D.bGLOBALJul 9, 2026

Meta to put AI chip into production in September as it looks to double computing capacity, memo shows

Meta Platforms plans to begin manufacturing its own AI chip in September 2026, targeting 14 gigawatts of total computing capacity by 2027, deepening compute centralization in one frontier AI firm.

WhyMeta proprietary chip production targeting 14GW total capacity deepens compute centralization in one frontier firm.Decentralized and federated architectures
RegressingMajorDesign · 7.D.bGLOBALJul 9, 2026

GPT-5.6 is now the preferred model in Microsoft 365 Copilot

GPT-5.6 was deployed as the preferred model in Microsoft 365 Copilot, deepening the integration between OpenAI and Microsoft across enterprise workflows and reinforcing incumbent concentration in frontier AI delivery.

WhyGPT-5.6 as default in M365 Copilot deepens OpenAI-Microsoft lock-in; only incumbents serve frontier models at this enterprise scale.Decentralized and federated architectures
AdvancingMajorLaws · 7.L.bEUROPEJul 8, 2026

European Court: Apple Can Not Shirk Off its Interoperability Requirements

The EU General Court ruled against Apple in multiple cases, affirming its Digital Markets Act interoperability obligations. The ruling prevents Apple from avoiding DMA compliance requirements, benefiting developers and users across Europe.

WhyEU General Court upholds DMA interoperability obligations vs Apple, advancing fair platform access for developers and users in Europe.Interoperability and data portability mandates
AdvancingMajorLaws · 7.L.aUSJul 6, 2026

Justice Department Reaches Proposed Settlement with Willow Bridge, One of America's Largest Landlords, to Resolve Information Sharing and Algorithmic Coordination Claims

The DOJ reached a proposed settlement with Willow Bridge, one of America's largest landlords, resolving claims about using algorithmic pricing systems to coordinate rental prices with competitors.

WhyDOJ proposes settlement with Willow Bridge over algorithmic rent coordination—conduct rules against AI-powered price collusion.Antitrust action against AI market concentration
AdvancingMajorLaws · 7.L.cEUROPEJun 19, 2026

Commission selects EUROPA consortium as the winner of the Frontier AI Grande Challenge, a project to build European open-source frontier AI model in all 24 EU languages

The European Commission has selected the EUROPA consortium as the winner of the Frontier AI Grande Challenge to build an open-source frontier AI model supporting all 24 EU languages.

WhyThe European Commission is funding the EUROPA consortium to build an open-source frontier AI model, advancing public-option AI.Public-option AI and sovereign compute funding