Surveillance Giants: How the Business Model of Google and Facebook Threatens Human Rights

Surveillance Giants: How the Business Model of Google and Facebook Threatens Human Rights is a report (index POL 30/1404/2019) published by Amnesty International on 21 November 2019, arguing that the surveillance-based business model of Google and Facebook — in which users access platform services by surrendering personal data that is monetised through targeted advertising — is structurally incompatible with the right to privacy and poses a systemic threat to the rights to freedom of opinion and expression, freedom of thought, and equality and non-discrimination. Secretary General Kumi Naidoo framed the report's core finding at launch: "Their insidious control of our digital lives undermines the very essence of privacy and is one of the defining human rights challenges of our era." The report is the first major publication from Amnesty International — one of the world's largest human-rights organisations — to treat surveillance capitalism as a structural human-rights problem rather than a compliance failure, and to call explicitly for a "radical transformation" of the two companies' core business model rather than incremental data-protection enforcement.

The Faustian bargain

The report's organising metaphor is a Faustian bargain: Google and Facebook provide services essential to contemporary digital life at no monetary cost, but users pay in the currency of intimate personal data — constantly tracked across the web and in physical space through connected devices — which is then fed into algorithmic systems designed to optimise targeted advertising and behavioural engagement. The report's human-rights framing turns on the non-voluntariness of this bargain: avoiding Google and Facebook is "simply not feasible" for most internet users, given the depth of each platform's integration into economic and social infrastructure. A user who refuses the terms gains access to no functional alternative; the market structure ensures that the rights-violating model is not a consumer preference but a structural condition of internet participation. The report's critique is therefore distinct from privacy law's typical consent-and-notice framework: consent that is extracted under conditions of no real choice is not consent that satisfies a human-rights standard, and enforcement of existing data-protection regimes under that framework will not resolve the structural incompatibility the report names.

Scale and market dominance

The report documents the scale at which the two platforms operate as the foundation for its claim of systemic threat. Combining users of its social platform, WhatsApp, Messenger, and Instagram, approximately one-third of all humans on Earth use a Facebook-owned service every day. Google accounts for approximately ninety percent of global internet search and operates Android on 2.5 billion monthly active devices. Together, the two companies collect roughly sixty percent of online advertising revenue and account for ninety percent of digital advertising market growth. The report frames this concentration not merely as an antitrust concern but as a human-rights architecture question: platforms of this scale, when their revenue model requires mass behavioural profiling, become surveillance systems at a population scale that no prior comparable infrastructure — state broadcasters, phone networks, postal systems — has operated. The algorithmic systems developed on top of that data infrastructure are designed not to serve users' stated preferences but to optimise for the companies' revenue interests, including through "behavioural nudges that keep people engaged on the platforms" and ad delivery tuned to exploit inferred psychological vulnerabilities.

Rights threatened and documented harms

The report identifies four rights as concurrently threatened by the surveillance model. The right to privacy is directly violated by mass data collection and the construction of detailed inferential profiles from behavioural data users have not explicitly disclosed. Freedom of opinion and expression is undermined by the combination of self-censorship produced by the knowledge of surveillance, algorithmic content moderation operating on training data biased toward majority populations, and the monetisation of attention through engagement-maximising content feeds that systematically amplify outrage and disinformation. Freedom of thought — the right not to have one's mental processes manipulated — is threatened by behavioural-influence techniques embedded in recommendation and targeting systems. Equality and non-discrimination is at risk from targeting systems that can deliver discriminatory outcomes — in employment, housing, and credit advertising — at population scale without the visibility required to detect or remedy them.

The report's primary documented-harm cases are Cambridge Analytica's misuse of Facebook user data and Google's unauthorised tracking of user geolocation — both instances where the underlying surveillance infrastructure enabled harms that the companies' own internal governance had not prevented and whose exposure required external investigation.

Recommendations

The report's recommendations address governments as its primary audience, reflecting a judgment that voluntary corporate reform is insufficient given the revenue incentives involved. The Amnesty campaign page structures the demands in three points. First, governments should ban surveillance advertising that relies on invasive tracking and the profiling of users for commercial targeting — the report frames this as the structural lever because surveillance advertising is what makes mass data collection economically rational in the first place; removing the revenue model removes the incentive. Second, governments should establish independent oversight of the algorithms behind the platforms to constrain the amplification of disinformation and harmful content that the engagement-maximisation model produces as a systemic byproduct. Third, governments should challenge Google and Facebook's market dominance to create conditions in which rights-respecting alternatives can be built and survive commercially — the report acknowledges that without structural market intervention, a hypothetical privacy-respecting entrant cannot compete at the scale required to be a real alternative. The press release frames the overall ask as a "radical overhaul" of the surveillance model rather than incremental regulation.

Position within the corpus

Surveillance Giants is the corpus's first Amnesty International publication entry, closing the surveillance-capitalism-from-human-rights-org slot in the publications layer. It arrived in November 2019 at the point when critique of the surveillance business model had achieved broad elite consensus — Apple's Tim Cook, various UN bodies, and now the world's largest international human-rights organisation on record — but regulatory responses remained unformed. The report's structural contribution is to translate that consensus from a principled position into a concrete recommendation architecture (the three-point government-action agenda) that subsequent campaign and regulatory advocacy could organise around. Amnesty's own campaign following the report formalised this trajectory, including work on the EU Digital Services Act consultation where the surveillance-advertising ban recommendation translated directly into the regulatory debate over targeted-advertising provisions. Within the publications layer the report sits alongside the Toronto Declaration (May 2018), which had established international human rights law as the binding normative register for machine-learning governance, and extends that framing from algorithmic systems specifically to the entire business model of the platforms on which those systems run — the argument is not that individual AI deployments violate human rights in isolable ways, but that the revenue logic that pays for and shapes the development of those systems is itself structurally incompatible with human rights. Structurally this is a significant escalation: where most 2018-2019 civil-society AI publications address deployment contexts (facial recognition, content moderation, hiring algorithms), Surveillance Giants attacks the financial substrate.