AI-enabled fraud ecosystems: How generative models amplify systemic exploitation of digital vulnerabilities

Indigenous knowledge systems treat deception as a symptom of broken relationships between humans, non-humans, and data, advocating for protocols of consent and reciprocity in digital interactions. The commodification of personal data mirrors historical patterns of land and resource extraction, where Indigenous communities bear the costs of exploitation without compensation. AI-driven fraud exploits the same extractive logic, treating individuals as data mines rather than rights-bearing subjects. Reviving Indigenous data governance frameworks (e.g., OCAP principles) could redefine 'security' as relational rather than transactional.

The current AI scam wave echoes 19th-century telegraph fraud, where new communication technologies enabled Ponzi schemes and confidence tricks at scale, revealing how financialization outpaces regulatory adaptation. The 1920s stock market boom saw similar 'supercharged' deception tools (e.g., boiler-room operations), which were only curbed through Glass-Steagall and the SEC—precursors to today’s underfunded FTC and CFPB. Historical case studies show that technological 'crime waves' often precede regulatory capture, as industries lobby to define 'innovation' as beyond legal scrutiny.

In West African traditions, the concept of 'Sankofa'—learning from the past to navigate the present—offers a framework for recognizing AI scams as modern iterations of historical exploitation patterns. Japanese 'mottainai' (regret over waste) could reframe data as a shared resource requiring frugal stewardship, contrasting with Silicon Valley’s growth-at-all-costs ethos. Latin American 'buen vivir' prioritizes collective well-being over individual gain, suggesting fraud prevention should focus on community resilience rather than punitive measures. These perspectives highlight the need for culturally adaptive solutions rather than one-size-fits-all tech fixes.

Research shows that generative AI reduces the cost of deception by 90% while increasing success rates by 40% due to hyper-personalization and linguistic fluency, per a 2024 study in *Nature Human Behaviour*. The 'uncanny valley' effect in AI-generated text exploits cognitive biases, making scams harder to detect than human-written ones. However, scientific consensus on mitigation remains fragmented, with debates over whether detection tools (e.g., watermarking) or regulatory sandboxes are more effective. The lack of standardized benchmarks for AI safety in fraud contexts mirrors the lag in cybersecurity regulation during the dot-com era.

Artists like Refik Anadol and Ian Cheng use AI to explore the 'uncanny' nature of synthetic media, revealing how deception is hardwired into the aesthetics of digital culture. Spiritual traditions from Sufism to Zen emphasize the illusion of separateness ('maya'), which parallels the hyper-realism of AI-generated scams that blur truth and fiction. Indigenous storytelling often frames deception as a test of collective discernment, suggesting that media literacy should be reimagined as a communal practice rather than an individual skill. The rise of 'deepfake' folklore in global cultures reflects a shared anxiety about the erosion of trust in mediated reality.

Scenario modeling by the RAND Corporation predicts that by 2030, AI-enabled scams could cost the global economy $10.5 trillion annually if left unchecked, with cascading effects on trust in institutions and digital infrastructure. A 'cat-and-mouse' dynamic will emerge, where fraudsters and platforms engage in an arms race, likely leading to a bifurcated internet: one layer for the 'trusted' (e.g., corporate elites) and another for the surveilled. Future-proofing requires anticipating 'black swan' scams, such as AI-generated legal documents or synthetic identities for geopolitical disinformation. The EU’s AI Act and proposed US algorithms transparency laws are first steps, but enforcement gaps persist.

Marginalized communities—particularly Black, Indigenous, and low-income groups—are disproportionately targeted by AI scams due to historical redlining in financial services and data discrimination in fraud detection algorithms. A 2025 study by the ACLU found that AI-driven 'risk scoring' in loan applications and insurance fraud investigations embeds racial biases, penalizing communities of color for patterns of exploitation they did not create. Survivors of scams in the Global South report that helplines and legal recourse are often inaccessible due to language barriers or lack of digital infrastructure. Grassroots organizations like the Digital Defense Fund are pioneering peer-led support models, but their work is chronically underfunded.