Are Open Hyperlinks Essential Infrastructure for Freedom of Thought?

Full report

On this page

The architecture of information discovery is undergoing a structural and epistemological shift. The transition from a decentralized, user-navigated web of documents to a centralized, machine-mediated oracle fundamentally alters how humanity accesses, interprets, and interacts with knowledge. This transition directly implicates the future of cognitive liberty, intellectual pluralism, and the freedom of scientific inquiry. As algorithmic gatekeepers and artificial intelligence (AI) models increasingly intermediate human access to knowledge, the hyperlink—a direct, unmediated vector between independent publishers—emerges not merely as a navigational tool, but as essential infrastructure for freedom of thought. The central inquiry of this multidisciplinary report evaluates the extent to which human-curated, independently operated hyperlink networks protect against concentrated corporate power, censorship, and algorithmic exclusion. Synthesizing computer science, information science, political philosophy, media studies, and civil-liberties scholarship, this analysis isolates the specific vulnerabilities of centralized discovery mechanisms. It further outlines technical, legal, and structural methodologies to rebuild a resilient web where the right to independent discovery is recognized and protected as a fundamental cognitive liberty.

1. Established Rights: The Legal and Philosophical Foundations of Cognitive Liberty

The debate over network architecture and hyperlink utility must be grounded in the established legal principles of intellectual privacy and the constitutional right to receive information, contrasting these rights against unresolved regulatory questions regarding digital intermediaries.

1.1. First Amendment Jurisprudence and the Right to Receive Information

In United States jurisprudence, it is an established right that the freedom of speech is inextricably linked to the corollary right to receive information. The Supreme Court formalized this duality in a series of landmark cases during the twentieth century. In Lamont v. Postmaster General (1965), the Court ruled that the federal government could not require citizens to affirmatively request the delivery of certain political mail, recognizing that placing administrative obstacles in the path of the receiver violated the First Amendment1. This doctrine was expanded in Stanley v. Georgia (1969), which affirmed that the "right to receive information and ideas, regardless of their social worth, is fundamental to our free society"1. The application of this right to curated information repositories was solidified in Board of Education v. Pico (1982). In Pico, the Supreme Court ruled that local school boards could not remove books from library shelves simply because they disagreed with the ideas contained within them2. The plurality opinion emphasized that the First Amendment protects the rights of students to receive a diversity of viewpoints, classifying the library as a mechanism of voluntary inquiry2. In the digital sphere, the open web functions as an analogous repository. When central intermediaries sever hyperlinks, suppress search results, or algorithmically demote visibility, they function as the de facto school board. The enclosure of independent publishing networks therefore threatens the legal spirit of the user's right to receive information, necessitating alternative architectures that protect independent discovery.

1.2. The Antimonopoly Tradition and Moody v. NetChoice

While the right to receive information is established, the regulatory classification of the digital platforms mediating this information remains a highly unresolved legal question. The 2024 Supreme Court case Moody v. NetChoice, alongside NetChoice v. Paxton, probes whether states can mandate viewpoint neutrality on large social media platforms7. Florida's S.B. 7072 and Texas's H.B. 20 sought to prevent platforms from algorithmically demoting or blocking users based on political ideology7. From an antimonopoly perspective, scholars argue that these platforms function as common carriers, historically subject to nondiscrimination requirements to protect the public's right to equitable access8. Conversely, technology companies argue that algorithmic recommendation and content moderation constitute "editorial discretion" protected by the First Amendment8. In Moody, the Court indicated that platform compilation and algorithmic curation do contain an expressive, editorial element, pushing back against the broad common-carrier designation7. However, if algorithmic recommendation is treated legally as protected editorial speech, user autonomy becomes entirely subordinate to the platform's opaque, value-laden judgments10. This legal reality emphasizes that the preservation of intellectual pluralism cannot rely on state-mandated neutrality within private algorithms. Instead, the structural solution must be rooted in decentralized protocols where users navigate independently via hyperlinks, bypassing the centralized editorial chokepoint altogether.

1.3. International Jurisdictions: The EU Digital Services Act

Regulatory approaches diverge significantly across international jurisdictions. Unlike the United States, where the First Amendment complicates the regulation of algorithmic curation, the European Union has established specific statutory limitations on recommender systems. Article 38 of the EU Digital Services Act (DSA) mandates that Very Large Online Platforms (VLOPs) and Very Large Online Search Engines (VLOSEs) offer users non-profiling alternatives in their recommender systems11. By forcing platforms to provide a chronological or strictly relevance-based feed stripped of behavioral profiling, the DSA attempts to restore a baseline of user agency11. While the DSA is a vital normative intervention, it relies on regulating centralized platforms rather than fundamentally decentralizing the architecture. A chronological feed within a walled garden still prevents the organic, cross-domain discoverability facilitated by open hyperlinks.

1.4. Intellectual Privacy and the Illinois Library Records Act

Freedom of thought requires intellectual privacy—the ability to read, explore, and formulate ideas free from surveillance or external judgment13. Before the digitization of information, this privacy was protected legally and structurally. For instance, the Illinois Library Records Confidentiality Act (75 ILCS 70/1) explicitly mandates that the registration and circulation records of a library are confidential, prohibiting their release to the public or law enforcement without a court order16. Modern digital discovery systems inherently violate this historical norm by requiring users to trade reading histories for algorithmic recommendations. Every click, dwell time metric, and query is ingested to build profiling models. The shift away from independently traversed hyperlinks toward centralized search engines and recommendation feeds has functionally eliminated intellectual privacy, making the adoption of zero-knowledge architectures a critical prerequisite for the restoration of cognitive liberty.

2. Cognitive Philosophy: The Web as an Extended Mind

The importance of the hyperlink extends beyond legal theory into the philosophy of cognition. In 1998, philosophers Andy Clark and David Chalmers proposed the "Extended Mind Theory" (EMT), arguing that cognitive processes do not terminate at the skull but extend into environmental artifacts used for thinking, memory, and calculation19. Under the framework of EMT, the web is not merely a tool accessed by the mind, but a constituent component of modern human cognition22. External memory offloading via the internet becomes a functional part of the human cognitive apparatus22. If the web operates as an extended mind, then the structural alteration of its topology by algorithmic intermediaries constitutes a form of cognitive interference. The suppression of independent hyperlinks—or their replacement by synthesized AI-generated summaries—severs the "synapses" of this externalized cognitive network. A user's inability to independently follow associative trails curtails their intellectual autonomy, effectively placing constraints on their capacity to think critically, synthetically, and expansively.

3. The Architecture of Information Discovery: Hyperlinks vs. Algorithms

The structural contrast between human-curated hyperlink networks and machine-curated feeds reveals profound differences in the patterns of viewpoint exposure, source diversity, and user autonomy.

3.1. Historical Precedents: Bush, Nelson, and the Ideology of Hypertext

The conceptualization of the hyperlink predates the internet and was explicitly designed to augment human thought. In 1945, Vannevar Bush described the "Memex," a theoretical mechanical device allowing users to store documents and link them together into "associative trails"24. Bush recognized that the human mind operates by association, rather than via rigid hierarchical indices25. Building upon Bush's theories, Ted Nelson coined the term "hypertext" in 1965 through Project Xanadu, advocating for "transclusion"—the seamless inclusion of a document into another, creating a highly interdependent, bi-directional network of ideas24. For Nelson, hypertext was not merely a technical specification; it was an ideological overhaul of literature, liberating readers from linear, top-down narratives and empowering individual choice in discovery28. In a true hyperlink network, the user operates as the spatial parser, and exposure to dissenting or unfamiliar ideas is dictated by the sociocultural entanglements of human authors who link out to sources, critics, and references27.

3.2. Network Topologies: The Bow-Tie Model

The empirical structure of these associative trails on the World Wide Web was definitively mapped by Broder et al. (2000) using the "Bow-Tie Model"29. The web graph consists of a core Strongly Connected Component (SCC) where pages link mutually to one another, representing a highly traversed intellectual commons30. Feeding into this core is the "IN" component consisting of pages that link to the SCC but are not linked back to, and an "OUT" component of pages linked to by the SCC but that do not link back30. Additionally, there are "tendrils" and "tubes" that bypass the center entirely30. Because the distribution of these links follows a power-law characteristic of scale-free networks (where a small number of nodes attract a massive proportion of links, governed by Zipf's Law), the discovery of niche information relies heavily on the long-tail connectivity of the graph29. Traditional search engines indexed this bow-tie and dropped users into various nodes across IN, SCC, and OUT, allowing the user to begin their associative journey.

3.3. Algorithmic Homogenization and Mixed-Grained Recommendation

Algorithmic recommendation feeds and LLM-mediated searches fundamentally disrupt this topology. Modern platform recommendation systems utilize Mixed-Grained Recommendation (MGR) models, which attempt to mathematically balance user "exploitation" (showing specific items a user is known to like) with "exploration" (introducing broader category-level information)32. However, these systems are ultimately tuned for platform retention and engagement, rather than intellectual pluralism or serendipitous discovery32. Empirical assessments demonstrate that personalization algorithms systematically distort learners' understanding of their environment. By over-representing a user's preferred categories and under-representing other dimensions, recommendation systems bias information sampling and induce incorrect generalizations34. In environments heavily mediated by LLMs, such as conversational search, users demonstrate a significantly higher tendency for selective exposure bias35. While a standard search engine returns a diverse list of links requiring user evaluation, LLM-powered interfaces synthesize a single, authoritative-sounding answer35. This synthesis frequently masks the diversity of underlying sources, obscures dissenting viewpoints, and enforces an artificially homogenized consensus.

4. Causal Diagram: Information Architecture, Source Diversity, and User Autonomy

To conceptualize the systemic consequences of centralized versus decentralized information architecture, the following matrix models the causal mechanisms linking technical design to human intellectual autonomy.

5. The Zero-Click Crisis and the Enclosure of the Web

The trajectory of centralized discovery is approaching an event horizon for independent publishers, driven by the rollout of AI Overviews and aggressive search suppression.

5.1. The Empirical Reality of AI-Mediated Gatekeeping

By early 2026, the proliferation of AI Overviews in search engines resulted in an environment where approximately 60% of all Google searches end without a single click to an external website36. Other metrics suggest that zero-click searches have reached as high as 68% of global queries38. The integration of AI summaries has reduced clicks to the number one organic search result by 58%36. The empirical impact on independent information distribution is devastating. Publishers reported a 38% year-over-year drop in Google referral traffic, while queries that trigger AI Overviews witnessed an organic click-through rate (CTR) crash of 61%36. Informational queries (e.g., "What is X?", "How to") are suffering an 85% and 72% zero-click rate, respectively36. The traffic that does survive the algorithmic filter is highly transactional; users who click through an AI overview exhibit 23x higher conversion rates due to their high intent, but the absolute volume of exploratory traffic has been functionally annihilated36.

5.2. Search Suppression and the Economic Starvation of Independent Publishing

In the context of the Bow-Tie model, the zero-click AI ecosystem acts as an information sink. The AI engine absorbs data from the Strongly Connected Component (SCC) and serves it directly on the search engine results page (SERP). Structurally, the user is prevented from entering the web's graph, terminating the associative trails originally envisioned by Bush and Nelson. By capturing the user at the point of query and refusing to distribute traffic to the originators of the information, centralized gatekeepers economically starve independent publishing. Without traffic, independent researchers, citizen journalists, and niche scientific publishers lose their ability to monetize, fundraise, or sustain operations. The AI-mediated web thus accelerates the hosting concentration of information into the hands of a few corporate technology giants, reducing the plurality of the intellectual ecosystem.

6. The Frailty of the Link: Link Rot, DNS Dependencies, and Centralization

If the open hyperlink is essential infrastructure for cognitive liberty, its physical and structural fragility poses a severe threat. Access to independent information must be resilient to accidental decay, structural dependencies, and intentional suppression.

6.1. The Ephemerality of the Web and the Zittrain Studies

The hyperlink is fundamentally ephemeral. A URL points to a server location, not a specific piece of immutable content, making it highly susceptible to "link rot" (the URL returns a 404 or 500 error) and "reference rot" or "content drift" (the URL returns a 200 success code, but the underlying content has fundamentally changed)39. Research spearheaded by Jonathan Zittrain at Harvard University demonstrated the catastrophic scale of this issue. A comprehensive study of U.S. Supreme Court opinions revealed that nearly 50% of hyperlinks cited in historical decisions pointed to content that had disappeared or drifted39. Similarly, a longitudinal study of the New York Times revealed that 53.5% of articles containing deep links featured at least one rotted link39. When the scholarly, journalistic, and legal foundations of a society rely on decaying infrastructure, historical accountability is severely compromised.

6.2. Infrastructural Solutions: Perma.cc, Memento, and Robust Links

To preserve the associative trails of the web, researchers and archivists have engineered protocols designed to robustify links and utilize distributed archives.

1. Perma.cc: Developed by the Harvard Law School Library Innovation Lab, Perma.cc allows authors to create a permanent, cached record of a cited webpage39. Librarians and scholars curate resources, saving a copy to the Perma servers. This ensures that legal and academic citations remain verifiable regardless of the origin server's future status40.

2. Memento Protocol (RFC 7089\): Memento adds a temporal dimension to HTTP. It allows a browser to request a prior version of a web page based on a specific datetime, seamlessly integrating current web infrastructure with distributed web archives like the Internet Archive and local library catalogs42.

3. Robust Links: Expanding on Memento, Robust Links decorate standard HTML hyperlinks with machine-actionable metadata42. A Robust Link includes the original resource URL (URI-R), the archival memento URL (URI-M), and the exact datetime the link was created43. If the original link rots or the content drifts, the browser can automatically degrade to the cryptographically verified archived version, ensuring uninterrupted access43.

6.3. DNS Dependencies and Mirror Sites

Even with robust archiving, the live web remains vulnerable to centralized Domain Name System (DNS) dependencies. The DNS hierarchy is governed by ICANN and localized registrars capable of revoking domains, functionally erasing nodes from the network. To counter this, independent publishers frequently rely on mirror sites—exact replicas of a primary site hosted across multiple geographic jurisdictions and top-level domains. By providing multiple hosting options and distributing access points, networks can route around localized censorship and DNS-level suppression.

7. The Decentralization Dilemma: Federated Publishing and its Discontents

Decentralization offers a powerful countermeasure against corporate gatekeeping, but it must not be assumed that decentralization automatically produces accuracy, fairness, or diversity. Decentralized architectures introduce unique threat vectors that require rigorous socio-technical modeling.

7.1. ActivityPub, Webmentions, and Bidirectional Discovery

The active distribution of independent links is shifting toward federated models. Standards like ActivityPub (a W3C recommendation) power the "Fediverse"—a decentralized network of independent servers (instances) that communicate seamlessly46. ActivityPub allows users to publish, follow, and interact across different domains without a central corporate authority48. Similarly, the Webmention protocol establishes a decentralized mechanism for websites to notify each other when they are linked46. Unlike proprietary comment sections or walled-garden ecosystems, Webmention allows independent blogs to build bidirectional relationships. A user on one server can reply to an article on another, and the original article can cryptographically verify and display that interaction49. This resurrects Nelson's vision of transclusion and bidirectional linking on an open, decentralized standard24.

7.2. Echo Chambers, Unsafe Directories, and Defederation

However, decentralized networks lack a central moderation authority, which can lead to the unchecked proliferation of coordinated disinformation, spam, and unsafe directories. While a centralized platform can universally ban a malicious actor or a botnet amplifying disinformation, a federated network relies on individual instance administrators to share blocklists, a process known as defederation48. This dynamic can unintentionally create strict ideological echo chambers. Entire instances frequently sever ties with one another over political or ideological disputes, fracturing the network into isolated sub-graphs governed by unaccountable curators. Consequently, a user on a highly moderated instance may be entirely shielded from outside viewpoints, inadvertently replicating the selective exposure bias of centralized algorithmic feeds.

8. Comparative Threat Model: Centralized vs. Decentralized Discovery Systems

To structure the security and sociological trade-offs between these paradigms, the following comparative threat model identifies the vulnerabilities of both centralized and decentralized discovery systems.

9. Privacy-Preserving Navigation: Metrics Without Surveillance

Freedom of inquiry requires the ability to navigate controversial, stigmatized, or highly sensitive information without the chilling effect of surveillance. However, independent publishers legitimately require analytics to understand navigation success, optimize site architecture, and detect denial-of-service fraud. Balancing these imperatives requires deploying zero-knowledge cryptographic architectures.

9.1. Oblivious HTTP and the Separation of Identity from Request

Oblivious HTTP (OHTTP), specified in RFC 9458, introduces a privacy-preserving mechanism that separates the identity of the user from the content of their request50. In standard network navigation, the host server logs both the client's IP address (identity/location) and the payload (the specific page requested). OHTTP fundamentally disrupts this by inserting an Oblivious Relay (OR) operated by a distinct third party50. The client encrypts the request using the target server's public key and sends it to the relay. The relay forwards the encrypted request to the target server but obscures the client's IP address. The target server decrypts and processes the request, but never knows who initiated the query50. Apple has successfully deployed this exact architecture in its Private Cloud Compute (PCC) infrastructure, ensuring that sensitive AI requests cannot be traced back to specific users while still allowing the system to verify authorization50.

9.2. Private Information Retrieval (PIR) and Computational Overheads

For deeper integration into discovery systems—such as querying a decentralized directory or fetching specific articles from an archive without the host knowing which article was read—Private Information Retrieval (PIR) must be utilized. PIR allows a client to retrieve an item from a database without the server learning which item was retrieved53. Historically, PIR was computationally prohibitive due to the overhead of processing the entire database for every query. However, recent advancements, such as the PIRacle system, leverage optimized Homomorphic Encryption (HE) to achieve massive efficiency gains. By optimizing homomorphic equality operators and reducing equality check overheads, modern PIR systems reduce server-side latency by up to 88.5x compared to legacy solutions53. By combining OHTTP for transport-layer anonymity and PIR for application-layer obfuscation, an independent discovery network can provide users with absolute intellectual privacy while still allowing network operators to measure aggregate bandwidth, total query volumes, and generic navigation success.

10. Technical Safeguards and Network Resilience

A decentralized, hyperlink-driven web relies fundamentally on trust. Without a central gatekeeper to filter out malicious actors, the network must rely on cryptographic verification of content and transparent governance models.

10.1. Source Provenance, C2PA, and Cryptographic Signatures

To combat deepfakes, AI hallucinations, and spoofing, the Coalition for Content Provenance and Authenticity (C2PA) established a technical specification for embedding cryptographic provenance data directly into digital assets54. A C2PA Manifest contains a digitally signed "Claim" detailing the asset's creation, the tools used (e.g., AI generation vs. optical camera), and the edits made over time54. The manifest is cryptographically bound to the content using hard hashes, allowing consumers to validate the authenticity of the information indefinitely54. However, cryptographic signatures are not a panacea. As the World Wide Web Consortium (W3C) has noted, while a digital signature prevents a malicious actor from inventing new, counterfeit data within an asset, it does not prevent an attacker from copying a genuine asset and its signature wholesale and hosting it in a deceptive context58. Therefore, C2PA manifests must be inextricably linked to the host domain's public keys. This ensures validation not just of the origin of the image or text, but the authorized context of its distribution.

10.2. Multiple Hosting Options, Transparent Governance, and Correction Procedures

Resilience requires that infrastructure be decoupled from singular physical locations. By employing decentralized storage networks (like IPFS) alongside traditional web hosting, content can be seeded across multiple hosting options, ensuring availability even if the primary domain is seized or censored. Furthermore, decentralized networks must establish transparent editorial governance and clear correction procedures. Unlike immutable blockchains where errors are permanently enshrined, the open web must utilize updateable pointers (such as decentralized identifiers, or DIDs) that allow authors to cryptographically sign corrections and amendments. These correction procedures preserve the scientific method of inquiry, ensuring that independent publishing remains self-correcting rather than dogmatic.

11. A Resilient, Privacy-Aware Hyperlink-Network Design

Synthesizing the technological building blocks analyzed in this report, the following architectural design is proposed for a resilient, privacy-aware hyperlink network resistant to AI-gatekeeping and corporate centralization. This serves as a foundational blueprint for CognitiveLiberties.com's research agenda:

1. Durable Identifiers & Content Addressing: Information is published using decentralized identifiers (DIDs) and hosted on content-addressable networks, decoupling the resource from strict DNS dependencies and geographic hosting vulnerabilities.

2. Cryptographic Provenance: All published text, data, and media embed C2PA manifests, signed by the author's private key, guaranteeing non-repudiation, transparent editing histories, and clear AI-disclosure54.

3. Robust Linking Protocol: All outbound hyperlinks format as Robust Links (URI-R, URI-M, Datetime). When an article is published, a Webmention triggers an immediate archival snapshot via the Memento Protocol (RFC 7089), ensuring the associative trail can never be broken by link rot43.

4. Federated Interaction: Discovery relies on open ActivityPub protocols, where users subscribe directly to publishers, library catalogs, or curated indices without algorithmic manipulation or MGR interference48.

5. Private Fetching: All requests for content or directory searches are routed through Oblivious HTTP relays and queried using Homomorphic Encryption via PIR. The publisher records a cryptographic proof of "navigation success" (a valid read) without capturing the reader's IP, device ID, or behavioral profile50.

12. Empirical Horizons: Five Testable Hypotheses

To transition the theoretical and legal analysis of link-driven discovery into empirical science, the following five testable hypotheses are proposed to guide future research on intellectual freedom:

1. The Autonomy Hypothesis: Users navigating information architectures based exclusively on decentralized, human-curated hyperlinks will demonstrate a statistically significant increase in exposure to cognitively distant, cross-ideological sources compared to users utilizing LLM-based conversational search, which homogenizes divergent views into a singular narrative.

2. The Archival Resilience Hypothesis: The systemic implementation of the Robust Links protocol and RFC 7089 (Memento) across a federated publishing network will reduce the long-term reference rot rate from the baseline of \~53% to under 5% over a longitudinal five-year study.

3. The Privacy-Exploration Hypothesis: Deploying Private Information Retrieval (PIR) and Oblivious HTTP (OHTTP) for content fetching will result in users demonstrating a higher willingness to explore stigmatized, dissenting, or politically sensitive topics, correlating an increase in intellectual privacy directly with broader information sampling.

4. The Decentralization Polarization Hypothesis: While federated architectures (e.g., ActivityPub) eliminate systemic single-point censorship, instance-level defederation policies will yield higher local network clustering (echo chambers) than the chronological feeds mandated by the DSA on centralized platforms.

5. The Zero-Click Deficit Hypothesis: As AI Overviews increase zero-click search rates beyond the 68% threshold, independent publishers relying on primary scientific or civic research will experience a disproportionate loss of exploratory audience traffic compared to transactional or commercial pages, fundamentally altering the economic viability of independent public-interest publishing.

13. Proposed Principles for a "Right to Independent Discovery" Charter

To enshrine these technical, legal, and philosophical imperatives into a normative framework suitable for public debate and policy development, this report proposes a Right to Independent Discovery Charter.

• Principle I: The Right to Unmediated Traversal. Every individual possesses the inherent right to access, follow, and traverse digital hyperlinks without covert algorithmic interception, rerouting, or artificial suppression by intermediary infrastructure providers.
• Principle II: The Guarantee of Intellectual Privacy. The act of reading and information discovery is a fundamental extension of human cognition. Users have an absolute right to zero-knowledge architectures. Analytics required for network maintenance must deploy cryptographic blinding (e.g., OHTTP, PIR) to permanently decouple human identity from reading histories.
• Principle III: The Imperative of Referential Integrity. A society's historical, scientific, and cultural record must remain intact. Publishers bear an ethical responsibility to deploy robust, machine-actionable archiving protocols (e.g., Memento, Perma.cc) to eradicate link rot and content drift, ensuring the continuity of the intellectual commons.
• Principle IV: The Sovereignty of Source Provenance. Information consumers have the right to cryptographic transparency regarding the origin, editing history, and AI-generation status of digital assets, independent of centralized platform fact-checkers or corporate arbiters of truth.
• Principle V: The Portability of Digital Identity. Users and publishers retain the right to own their social graphs and discovery networks. The ability to port cryptographic identities, follower networks, and content across hosting environments is essential to resist digital monopolization and censorship.

Conclusion

The architecture of the internet is not merely a technical configuration; it operates as the structural scaffolding of modern human cognition. The open hyperlink represents a vital mechanism for cognitive liberty, facilitating self-directed associative trails across an expansive diversity of human thought. As zero-click ecosystems, large language models, and opaque algorithmic recommenders threaten to collapse the web's associative trails into highly managed, centralized walled gardens, the open hyperlink remains the primary defense for intellectual pluralism. Rebuilding a stronger, link-driven web—fortified by robust archiving, cryptographic provenance, decentralized governance, and zero-knowledge privacy protocols—is not an exercise in digital nostalgia. It is an urgent civil liberties imperative. To protect the freedom of thought, scientific inquiry, and democratic pluralism in the twenty-first century, society must vigorously defend, regulate for, and technically guarantee the unalienable right to independent discovery.

Works cited

1. Art by Telephone, Art by Algorithm: Expression, AI, and the First, https\://cdt.org/insights/art-by-telephone-art-by-algorithm-expression-ai-and-the-first-amendment/

2. The Constitutionality of the Children's Internet Protection Act, https\://scholarship.stu.edu/cgi/viewcontent.cgi?article=1534\&context=stlr

3. The Right to Receive Information: The Current State of the Doctrine, https\://scholarship.shu.edu/cgi/viewcontent.cgi?referer=\&httpsredir=1\&article=1210\&context=shlr

4. The National Coalition Against Censorship \- Supreme Court, https\://www\.supremecourt.gov/DocketPDF/25/25-284/379559/20251014141805409\_No.%2025-284\_Amici%20Brief.pdf

5. BOARD OF EDUCATION v. PICO, 457 U.S. 853 (1982) | FindLaw, https\://caselaw.findlaw.com/court/us-supreme-court/457/853.html

6. Banned Books: Censorship and Protecting Students from, https\://www\.okbar.org/banned-books-censorship-and-protecting-students-from-censored/

7. The First Amendment and Platform Content Moderation, https\://www\.internetgovernance.org/2024/07/08/the-first-amendment-and-platform-content-moderation-the-supreme-courts-moody-decision/

8. Teachout.pdf \- Temple Law Review, https\://www\.templelawreview.org/lawreview/assets/uploads/2024/07/Teachout.pdf

9. Online Speech at the US Supreme Court in Moody v. Netchoice, https\://verfassungsblog.de/online-speech-at-the-us-supreme-court-in-moody-v-netchoice/

10. Federal Court Rules That Blocking Speech Is Protected By The First, https\://www\.techdirt.com/2026/08/17/federal-court-rules-that-blocking-speech-is-protected-by-the-first-amendment-but-recommending-it-isnt/

11. Understand the Connection Between DSA and GDPR \- EDAA, https\://edaa.eu/digital-services-act/dsa-and-gdpr/

12. How the Digital Services Act Addresses Platform Recommender, https\://www\.mozillafoundation.org/en/blog/action-recommended-how-the-digital-services-act-addresses-platform-recommender-systems/

13. BIPA Reforms Amplify Need for Federal Standards in Biometric Data, https\://scholarship.law.umn.edu/cgi/viewcontent.cgi?article=1578\&context=mjlst

14. Stay Out of My Head: Neurodata, Privacy, and the First Amendment, https\://scholarlycommons.law.wlu.edu/cgi/viewcontent.cgi?article=4839\&context=wlulr

15. Recording as Heckling \- Georgetown Law, https\://www\.law.georgetown.edu/georgetown-law-journal/wp-content/uploads/sites/26/2019/11/Recording-as-Heckling.pdf

16. Standards for Illinois Public Libraries, https\://www\.lislelibrary.org/sites/default/files/assets/Policies/Standards%20for%20Illinois%20Public%20Libraries%203\_0.pdf

17. (75 ILCS 70/) Library Records Confidentiality Act. \- ILGA.gov, https\://www\.ilga.gov/Legislation/ILCS/Articles?ActID=1004\&ChapterID=16\&Print=True

18. Record Confidentiality Policy \- Prairie Skies Public Library District, https\://www\.pspld.com/record-confidentiality-policy

19. The beauty of research opens up a whole world of opportunity. The, https\://www\.ibsafoundation.org/en/blog/extended-mind-the-beauty-of-research

20. I am having trouble understanding whether Extended Mind has merits, https\://www\.reddit.com/r/askphilosophy/comments/kftjf1/i\_am\_having\_trouble\_understanding\_whether/

21. the cognitive turn of the extended mind theory” 25 Anos da Tese da, https\://www\.scielo.br/j/trans/a/sVbqRqRMsfhr4DrDqxJ4s3v/

22. Extended Cognition and Functionalism \- ResearchGate, https\://www\.researchgate.net/publication/228680785\_Extended\_Cognition\_and\_Functionalism

23. the cognitive turn of the extended mind theory” 25 Anos da Tese da, https\://www\.redalyc.org/journal/3842/384275142014/

24. What comes after Roam's renaissance? \- LessWrong, https\://www\.lesswrong.com/posts/CoqFpaorNHsWxRzvz/what-comes-after-roam-s-renaissance

25. The Brown/MIT Vannevar Bush Symposium, https\://cs.brown.edu/memex/Bush\_Symposium\_Interact\_2.html

26. Each one of us was several: networks, rhizomes and web organisms, https\://repository.londonmet.ac.uk/3804/3/Tredinnick-KOJ-final-revised.pdf

27. Seven Hypertexts \- ePrints Soton \- University of Southampton, https\://eprints.soton.ac.uk/482482/1/HT\_2023.pdf

28. The Magical Place of Literary Memory™: Xanadu \- Screening the Past, https\://www\.screeningthepast.com/issue-18-first-release/the-magical-place-of-literary-memory%E2%84%A2-xanadu/

29. Characterization of National Web Domains \- Harding University, https\://sites.harding.edu/fmccown/classes/archive/comp475-s09/char-national-web-domains.pdf

30. A complex model for European Union legal sources \- arXiv, https\://arxiv.org/pdf/1501.05237

31. Essays on Applied Network Theory \- IRIS Unimore, https\://iris.unimore.it/bitstream/11380/1374315/2/Thesis\_Flori.pdf

32. A Novel Deep Learning Method for Mixed-Grained Recommendation1, https\://misq.umn.edu/misq/article/50/3/1117/3762/Deep-Chain-of-Preference-A-Novel-Deep-Learning

33. A factual and perceptional framework for assessing diversity effects, https\://www\.emerald.com/intr/article/29/6/1526/185574/A-factual-and-perceptional-framework-for-assessing

34. Algorithmic personalization of information can cause inaccurate, https\://pmc.ncbi.nlm.nih.gov/articles/PMC12353355/

35. Generative Echo Chamber? Effects of LLM-Powered Search ... \- arXiv, https\://arxiv.org/html/2402.05880v2

36. 60% Zero-Click Searches: The 2026 SEO Crisis Strategy, https\://www\.digitalapplied.com/blog/60-percent-searches-zero-click-crisis-2026-seo-strategy

37. AI Answers Are the \#1 Cause of Traffic Decline in 2026 \- xSeek, https\://www\.xseek.io/blogs/articles/ai-traffic-decline-2026

38. Zero-click searches hit 68% — what it means for your business, https\://www\.facebook.com/quartznews/posts/zero-click-searches-hit-68-what-it-means-for-your-business-zero-click-searches-n/1460158335979980/

39. An Examination of Linkrot and Content Drift within The New York, https\://dash.harvard.edu/bitstreams/10737af1-7d3d-459b-8ef8-177a6b27add8/download

40. Perma: Scoping and Addressing the Problem of Link and Reference, https\://harvardlawreview.org/forum/vol-127/perma-scoping-and-addressing-the-problem-of-link-and-reference-rot-in-legal-citations/

41. Pausing the Internet \- Harvard Law School Center on the Legal, https\://clp.law.harvard.edu/article/pausing-the-internet/

42. Memento \- Time Travel for the Web, https\://mementoweb.org/about/

43. Robustifying Links To Combat Reference Rot \- The Code4Lib Journal, https\://journal.code4lib.org/articles/15509

44. Supporting Account-based Queries for Archived Instagram Posts, https\://digitalcommons.odu.edu/context/computerscience\_etds/article/1139/viewcontent/Jayanetti\_thesis.pdf

45. Examining persistence of European open repository infrastructure, https\://arxiv.org/pdf/2601.04015

46. Micropub \- Indie Microblogging, https\://book.micro.blog/micropub/

47. Webmention Explained: How Websites Notify Each Other When, https\://dallol.observer/2026/09/09/webmention-explained-how-websites-notify-each-other-when-they-are-linked/

48. The Social Web Explained: How ActivityPub Powers the Fediverse, https\://www\.technodabbler.com/the-social-web-explained-how-activitypub-powers-the-fediverse/

49. Microformats \- Indie Microblogging, https\://book.micro.blog/microformats/

50. Request Flow | Documentation \- Apple Security Research, https\://security.apple.com/documentation/private-cloud-compute/requestflow

51. privacy-routing.md · pear-git-v3 \- 7rnx, https\://7rnx.net/git/repos/pearbrowser-desktop/blob/docs/research/privacy-routing.md

52. shade-tree-node/docs/residential-proxies.md at main · dmarzzz, https\://github.com/dmarzzz/reputation-gated-onion-egress/blob/main/docs/residential-proxies.md

53. PIRacle: A Fast and Scalable Private Information Retrieval System, https\://www\.computer.org/csdl/journal/tc/2026/04/11359596/2dsHojHBHUs

54. Content Credentials : C2PA Technical Specification, https\://spec.c2pa.org/specifications/specifications/2.0/specs/C2PA\_Specification.html

55. C2PA Implementation Guidance, https\://spec.c2pa.org/specifications/specifications/1.2/guidance/Guidance.html

56. C2PA Technical Specification, https\://spec.c2pa.org/specifications/specifications/1.0/specs/C2PA\_Specification.html

57. Content Credentials : C2PA Technical Specification, https\://spec.c2pa.org/specifications/specifications/2.4/specs/C2PA\_Specification.html

58. W3C pushes revocable ID barcodes after 160 million ... \- PPC Land, https\://ppc.land/w3c-pushes-revocable-id-barcodes-after-160-million-drivers-licenses-leak/