You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
The US federal government manages more than 1/4 of US land, encompassing large wildlife populations. Government decisions concerning these lands can affect wild animals at scale. The government has started to use AI for permitting and wildlife management decisions, but obliquely and with few sideboards. If AI systems become embedded in decisions affecting wild animals, while the animals are represented only as regulatory variables (population counts, habitat acreage, extinction thresholds, etc), there is a risk that the devaluation of wild animal welfare also become entrenched. To prevent the acceleration and lock-in of decision-making processes that could effect enormous numbers of wild animals, it is necessary to identify how AI-assisted decision making can affect decisions that determine the treatment and survival of individual wild animals at scale and how to build animal interests into those systems before they become entrenched.
Identify legally significant patterns in government use of AI in land management decision-making that affects wild animal populations and assess opportunities to ensure animal welfare.
The money will be used to fund my work to identify agency practices, request and review government records, develop best legal theories, compare AI-assisted analysis with conventional review, and identify intervention points (at 50% time over the next six months).
Right now, it's just me. I have spent decades identifying how government agency decisions translate into real world effects on wildlife, how to engage in the administrative process early enough to influence the outcomes of the decisions, how to obtain information when companies and agencies refuse to disclose it, and how to challenge the government's decisions in court, when necessary. In the past six months I have begun researching and tracking agency use of AI in environmental decision making and developing legal theories about the government's obligations regarding its use of AI.
Failure would be the inability to get any more information about how the government is using AI, and the cause would be the government's refusal to provide information as required by the law. The result would be an opportunity to challenge that decision in court.
I haven't raised any for this project, yet.