Algorithmic Accountability Act of 2025 (H.R.5511): requiring impact assessments for automated decision systems

1Automated decisions that shape daily life

In important moments of our lives, whether we can borrow money, get hired, rent a home, or obtain insurance, decisions are increasingly made by algorithms and AI rather than people. Automation is efficient, but if it inherits the biases in its training data, it can treat certain people unfairly. And when the reasons for a decision are opaque, those affected find it hard to understand or seek redress.

Against this spread of automated decision-making, the idea of holding such systems accountable for their impacts is algorithmic accountability.

2Requiring impact assessments for automated decisions

H.R.5511, the Algorithmic Accountability Act of 2025, is a House bill that seeks to put this idea into law. As its name indicates, it appears aimed at requiring accountability for the impacts of automated decision systems. The sponsor is Rep. Yvette Clarke (D-NY), and with 30 cosponsors as of 2026-07-05 it shows notable interest in the theme.

It was introduced on September 19, 2025, and referred to the House Committee on Energy and Commerce. Which systems would be covered, who would perform what kind of assessment (such as impact assessments), and where it would be reported are provision details beyond the information here, so no firm conclusions are drawn.

3When a person decides, and when a system does

Whether you can borrow money, be hired, rent a home, obtain insurance. When the decider in those moments shifts from a person to a system, what changes lines up as follows.

Algorithmic accountability is framed as the response to that difference. The bill was introduced on 19 September 2025 and referred to the House Committee on Energy and Commerce, with 30 cosponsors as of 2026-07-05. Which systems are covered, who performs what assessment and where it is reported are details beyond what is shown here, and this site does not assert them.