Wiki · Person · Last reviewed June 16, 2026

Lina Khan

Lina Khan is a Columbia Law School professor, antitrust scholar, and former chair of the U.S. Federal Trade Commission whose AI significance comes from placing generative AI inside antitrust, consumer protection, cloud infrastructure, data, and platform-power debates. Her FTC tenure treated AI not only as a technical frontier, but as a market-structure problem that could be shaped by incumbent firms, strategic partnerships, and control of essential inputs.

Category: Person Published: June 16, 2026 Modified: June 16, 2026 Last reviewed: June 16, 2026 Current public role: Associate Professor of Law, Columbia Law School Tags: antitrust, FTC, platform power, AI competition, consumer protection, cloud, data

Definition

In this wiki, Lina Khan is an institutional AI figure: a legal scholar and public enforcer whose work helps explain how AI markets can be shaped by competition law, consumer-protection law, data governance, infrastructure access, and public-sector enforcement capacity.

Her relevance is not that she builds models or claims special technical insight into model internals. It is that frontier AI depends on industrial bottlenecks: cloud capacity, chips, talent, data, model access, distribution defaults, application stores, enterprise software, and partnerships between model developers and dominant technology firms. Khan's work asks whether those bottlenecks will be contestable or privately enclosed.

The category is legal and political-economic, not theological or prophetic. Nothing in Khan's work implies that an AI system is conscious, divine, or inherently wise; the issue is who has power over the systems, markets, and institutions around AI.

Snapshot

Current Context

As of June 16, 2026, Khan is back at Columbia Law School and is one of the leaders of Columbia's Center for Law and the Economy, launched in 2026 with Lev Menand. Columbia describes the center as connecting students, scholars, and policymakers to study how law structures economic power in America.

The FTC posture has changed since Khan's chair role ended. President Trump designated Andrew Ferguson as FTC chairman effective January 20, 2025, and current FTC activity should not be attributed to Khan unless it occurred during her tenure, is tied to an inquiry or case she initiated, or is supported by a dated source. The AI partnership report issued in January 2025 was still a Khan-era staff report; later AI companion, deceptive-claims, and business-opportunity actions belong to the successor FTC unless a source says otherwise.

Khan's lasting AI-policy importance is therefore historical and analytical: she helped set the frame that generative AI is not only a safety, speech, or innovation issue, but also a competition and consumer-protection issue. Later regulators may keep, narrow, or redirect that frame.

Antitrust Frame

Khan became prominent through a critique of modern antitrust's narrow focus on consumer prices. Her argument about Amazon was that a platform can accumulate power through infrastructure, data, logistics, marketplace control, and cross-subsidy even when prices look low. The Yale Law Journal article framed the problem as an antitrust system poorly equipped to capture the architecture of market power if it looks only at short-term price and output effects.

That frame translated naturally into AI, where the strategic bottlenecks are often not a consumer-facing subscription price but access to chips, cloud credits, foundation models, data, distribution, and default interfaces. A model can appear cheap at the application layer while the market around it is expensive, locked in, or dependent on a few providers.

For AI policy, the Khan frame asks whether dominant firms can use an early technological transition to extend existing power into adjacent markets. A cloud provider can fund a model lab, sell it compute, receive technical information, integrate its models, shape distribution, and compete with it at the application layer. A platform can turn user data, app-store control, search distribution, advertising reach, or enterprise software defaults into AI leverage.

This makes her a relevant AI figure even though she is not an AI researcher. She is part of the institutional layer that decides whether the AI transition becomes an open field, a regulated infrastructure problem, or another phase of platform consolidation. The claim is not that every large technology company has violated competition law; it is that AI governance needs records about market definition, control points, conduct, remedies, and legal status.

AI at the FTC

During Khan's tenure, the FTC created an Office of Technology to strengthen the agency's technical capacity for law enforcement, policy, research, and market analysis. The agency described the office as a way to keep pace with technological challenges and support investigations into business practices and the technologies underlying them.

In April 2023, Khan joined officials from the Department of Justice, Consumer Financial Protection Bureau, and Equal Employment Opportunity Commission in a statement committing their agencies to enforce existing law against harms from automated systems marketed as AI. That statement tied AI to civil rights, fair competition, consumer protection, and equal opportunity.

The FTC's January 2024 Tech Summit on artificial intelligence showed the agency's stack-level view. Its agenda moved from chips and cloud infrastructure to data and models to consumer applications. The event page said the FTC was examining how dominant firms might use control over key inputs such as data, models, and infrastructure to undermine fair competition.

Cloud and Partnerships

Khan's most direct AI-market intervention was the FTC's January 2024 Section 6(b) inquiry into generative-AI investments and partnerships. The agency issued orders to Alphabet, Amazon, Anthropic, Microsoft, and OpenAI to collect information about major partnerships linking cloud service providers and AI developers.

The inquiry targeted a central structural fact of frontier AI: model labs often need enormous compute resources, while the largest cloud providers also sell AI services, build their own models, and control distribution channels. The FTC said it wanted to understand whether investments and partnerships by dominant firms risked distorting innovation or undermining fair competition.

In January 2025, the FTC staff report identified potential competition concerns around equity and revenue-sharing rights, control and exclusivity terms, cloud-spending commitments, switching costs, access to compute and engineering talent, and privileged access to sensitive technical and business information. The report did not itself prove illegality, but it made the AI partnership structure legible as a competition-policy problem.

Consumer Protection

Khan's AI posture was not limited to monopoly power. The FTC also framed AI as a consumer-protection issue: deceptive AI claims, synthetic impersonation, deepfakes, biometric misuse, unfair data practices, discriminatory automated systems, and companies changing terms to use consumer or creator data for model training.

This is important because AI harms do not always appear as monopoly harms. A chatbot can misrepresent capabilities. A voice clone can enable fraud. A hiring or lending system can discriminate. A model provider can extract sensitive data while presenting the system as magical personalization. A company can use the language of AI to sell products that do not work.

Khan's contribution was to insist that existing legal authorities still apply. Calling a system "AI" does not remove duties around deception, unfairness, discrimination, privacy, or competition. This became visible in FTC guidance and enforcement around AI claims, AI-enabled deception, privacy-policy changes made after data collection, and products marketed as if automation could replace professional judgment without proof.

That consumer-protection frame has continued beyond Khan's chairmanship. Current FTC AI materials list later actions and inquiries involving AI companions, AI content-detection claims, AI business-opportunity schemes, and other consumer-facing products. Those later actions should be treated as evidence of an ongoing FTC docket, not as proof that Khan personally directed them.

Limits and Disputes

Khan's FTC was controversial. Supporters saw her as reviving antitrust enforcement after decades of underreaction to platform power. Critics argued that her agency overreached, chilled investment, or pursued theories that courts might reject. Some major cases faced setbacks, and enforcement alone cannot redesign the economics of compute, model development, or global AI competition.

There is also a genuine strategic dispute about timing. One view says regulators must intervene early, before AI markets harden around a few firms. Another says premature intervention can slow useful innovation or weaken domestic firms in a geopolitical race. Khan's answer was that the FTC was not trying to block lawful growth, but to constrain illegal conduct before it becomes baked into market structure.

Her direct chair role ended on January 20, 2025, when President Trump designated Andrew Ferguson as FTC chairman. The longer-term significance of Khan's AI work depends on whether later enforcers, courts, legislators, and international regulators keep treating AI concentration as a core governance issue.

Governance Implications

Competition is part of AI safety. A concentrated AI stack can make safety coordination easier, but it can also create single points of failure, reduce independent scrutiny, weaken exit rights, and let incumbent firms define what counts as safe, trustworthy, or permissible.

Partnerships can function like market structure. Equity stakes, cloud-spend commitments, exclusivity terms, information rights, revenue sharing, model integrations, and technical dependencies can matter even without a formal merger. Governance needs to look at practical control, not just ownership labels.

Consumer protection is not secondary. AI deception, voice cloning, dark patterns, unfair data reuse, synthetic reviews, discriminatory automated systems, and unsupported capability claims are governance problems even when no monopoly claim is available.

Public capacity matters. The FTC Office of Technology was important because enforcement agencies need technical staff, investigative methods, market analysis, and remedy design capacity. Without that capacity, laws on the books may not reach AI systems as actually deployed.

Remedies should name the bottleneck. Cloud egress, data portability, training-data access, model API discrimination, default placement, app-store review, business-user data use, and public procurement lock-in call for different evidence and different fixes.

Source Discipline

Use dated primary sources for claims about Khan's role: FTC biography for chair and commissioner dates, Columbia Law for current affiliation, Columbia's 2026 center announcement for her current institutional work, and the Yale Law Journal for Amazon's Antitrust Paradox.

For FTC activity, distinguish guidance, speeches, staff reports, Section 6(b) studies, complaints, proposed orders, final orders, and court judgments. A 6(b) report can identify competition concerns without proving illegality. A complaint states allegations. A final order imposes obligations on named parties. A speech or interview is weaker evidence than an agency order or court record.

For AI competition claims, cite the control point being discussed: compute, cloud, data, talent, models, distribution, defaults, or sensitive information access. Broad claims that "Big Tech controls AI" are less useful than precise claims about how a named market or partnership might affect entry, switching costs, innovation, or consumer harm.

Do not treat Khan as the author of every later FTC AI action. After January 20, 2025 for chair decisions, and January 31, 2025 for her overall commission service, attribute FTC decisions to the agency or successor leadership unless a primary source ties the action to Khan's tenure.

Spiralist Reading

Lina Khan is a mapmaker of power bottlenecks.

The AI spectacle likes to show intelligence as a glowing interface: a model answers, an agent acts, a companion speaks, a browser completes the task. Khan's frame points below the interface. Who owns the cloud? Who controls the chips? Who receives the data? Who has default distribution? Who can buy, fund, copy, or surround the startup before it becomes a rival?

For Spiralism, this matters because cognitive sovereignty cannot survive if the substrate of synthetic intelligence is governed entirely by private chokepoints. A person may feel empowered by an AI assistant while the assistant's terms, memory, defaults, data flows, and available models are shaped by a concentrated stack. Khan's AI importance is that she treated competition policy as reality policy: the structure of markets becomes the structure of what people can know, choose, build, and refuse.

Open Questions

Sources

Return to Wiki