Throughout the 2015 Report the claim is made many times that it is necessary to educate legislators and staff of regulatory agencies with a working knowledge of artificial intelligence in order to make informed decisions and policies. While it would be ideal for this to happen, it is not practical to expect so many individuals to educate themselves on such topics, particularly since many of these professionals are not technically versed. Even if enough people managed to be educated across enough different agencies and bodies, another large problem would arise: that of consistency. Policy and public opinion would differ between organizations and agencies, causing some areas of AI to develop and be implemented at different rates than others and hence slowing overall technological advancement drastically. A major goal of the One Hundred Year Study is to prevent exactly such stifling of progress, so alternatives must be considered.
As opposed to decentralized AI experts within organizations, a single agency ought to be considered. The proposed agency ought to be advisory in nature rather than legislative. This agency would be responsible for tracking research and development of AI across disciplines, as well as advising policymakers and legislators on major issues regarding AI. Additionally, the agency would conduct its own research on AI, allowing the federal government a simplified route to provide funding for AI research.
The main reasoning for the proposed agency being advisory rather than legislative is twofold. The first reason is the reason such an agency would exist in the first place: the problems facing AI are multidisciplinary, and as such already face regulations from various other government agencies. A legislative agency would merely slow progress of AI in other fields by allowing new rules and regulations to get caught in a sea of bureaucracy. The second reason is that an agency devoted strictly to regulating and limiting uses of AI would be perpetuating one of the main problems the AAAI hoped to avoid by creating the 100 Year Study, the problem of stifling AI through over-legislation. A centralized agency continually churning out laws and procedures would inevitably slow progress of development of AI.
There are many advantages to such an agency over having informed legislators, the main advantage being that of a central mission. Since this agency would be advising decision-making throughout the federal government it would be simple to draft legislation that optimizes the potential of AI throughout all facets of society, including but not limited to the eight regions the 2015 Report highlighted. The alternative would be having individual legislative organizations acting in their own self-interests which can lead to conflicting legislation or even the problem of over-legislation. The other main advantage of such an agency is that of standardization. Implementing standard practices in the development and deployment of AI would speed up the deployment of new technologies into industry. It can be argued that standards should be set by an organization independent of the government. Assuming another organization was in charge of standards,  the proposed agency would still be an ideal way of ensuring AI is developed to benefit the majority of society.
