GiveDirectly and Google.org are using AI mapping and satellite imagery to find low-income hurricane survivors in Puerto Rico and Florida and send cash payments
Context & Ripple Effects
This program extends a decade-long Google.org pattern of funding AI-for-social-good grantees, from the $25M AI impact grants of 2019 to the economic-equity grants for women and girls. What changed by late 2022 is the mechanism: instead of funding research, Google's satellite and mapping stack — including Earth Engine's imagery and analysis tools — is now the targeting engine for direct cash transfers.
GiveDirectly's role matters here: the nonprofit built its model on identifying poor households and sending unconditional payments, and pairing it with Google's imagery lets it find survivors rather than wait for applications. The related coverage shows where this leads — by 2025 the same pairing moves from disaster response to disaster anticipation.
First-order effects
- Low-income households in Puerto Rico and Florida receive cash payments based on satellite-detected damage and income proxies, bypassing application-based aid queues entirely.
- GiveDirectly's targeting cost structure shifts: AI mapping replaces manual needs assessment as the primary way it decides who gets paid after a hurricane.
Second-order effects
- Other disaster funders and nonprofits face a new benchmark for speed and precision — if Google's imagery can identify recipients, slow survey-based disbursement becomes harder to defend.
- Google's philanthropic AI tools become a distribution advantage: Earth Engine and its forecasting models turn Google infrastructure into the default backend for humanitarian targeting, deepening nonprofit dependence on one company's data stack.
Third-order effects
- The trajectory runs from reactive to anticipatory aid: Google Research's flood forecasting across 80+ countries and the planned Flood Hub cash-transfer pilot in Bangladesh point toward aid dispatched before impact, not after.
- If satellite-targeted transfers prove out across disasters, humanitarian response consolidates around platforms that own both the imagery and the payment rails — raising questions about who audits the algorithms deciding who counts as needy.
The trend: Humanitarian aid is shifting from application-based relief to AI-targeted cash transfers, with tech-company geospatial platforms becoming the targeting layer and disaster prediction the next frontier.