The risks

Data and the cloud concentrate risks that are less visible than those of AI or blockchain, but more systemic, because they support the entire digital economy.

Three American companies carry more than 60% of the world's cloud infrastructure. When one of them sneezes, the world catches a cold. At company level, the risk takes the form of vendor lock-in: proprietary services, egress fees, migration costs.

The US CLOUD Act allows US authorities to demand access to data from American providers, wherever it is hosted. EU-US frameworks have repeatedly been challenged, which turns legal risk into strategic risk.

Personal data is the raw material of an economy of commercial surveillance: data brokers, targeted advertising, profiling. The GDPR gave Europe a protective framework, but its enforcement and its balance with innovation are still debated.

Data is attackers' number one target. A cloud specificity: many leaks come from a customer configuration error, an illustration of a poorly understood shared responsibility model.

Data centres consume a growing share of the world's electricity, driven by AI. Location changes everything: a data centre in France emits far less CO2 than one powered by very carbon-intensive electricity.

Incomplete or biased data produces biased decisions, and biased AI. Now that data feeds the models, its blind spots become theirs.