Google Permits High-Risk AI Decisions Under Human Supervision
Google updates its generative AI policy to explicitly allow automated decision-making in high-risk domains like healthcare and housing, provided a human supervises the process.
Google updates its Generative AI Prohibited Use Policy to explicitly allow customers to deploy its AI tools for automated decision-making in high-risk domains. Provided a human supervisor remains in the loop, organizations can now use Google's generative AI to make decisions that potentially impact individual rights in areas such as healthcare, employment, housing, insurance, and social welfare.
A Google spokesperson clarifies that the human supervision requirement always existed in the company's policy, but the latest update simply recategorizes items and provides clearer examples for users. The previous wording of Google's terms misleadingly implies a blanket ban on high-risk automated decision making, even though the company insists this practice is permitted as long as a human oversees the process.
This updated stance sets Google apart from its top AI rivals, which enforce much stricter rules. OpenAI completely prohibits the use of its services for automated decisions relating to credit, employment, and housing, while Anthropic requires both a qualified professional's supervision and explicit customer disclosure. These policy differences emerge as regulators and advocacy groups increasingly scrutinize AI decision-making tools for perpetuating historical discrimination and bias.