The Need for AI Public Policy Innovation
A criminal investigation is underway in Miami of ChatGPT (OpenAI) being responsible for murder.
We have already seen isolated violent incidents. Others are predicting street demonstrations in the next few months. Some say these are just coming from people with mental problems. That may be true. But, are the mentally disturbed just the canaries in the coal mine?
The only way out of this is the creation of public policy measures such that a rising tide lifts all boats. That is, all aspects of society will benefit from AI in roughly equivalent fashion. How can we achieve this?
Adaptation Through Regulation Background
There is a group of AI investors and companies that have called for a laissez-faire approach to AI. These people have argued that any public policy that involves regulation will slow down the necessary innovation and therefore should be avoided at all costs. This group has lobbied in the US and created a political action committee that funds politicians who support their position. This group was successful in getting the Trump administration to support their approach.
The US Department of Defense (DoD) has sanctioned Anthropic in an attempt to force the company to allow the DoD to use Anthropic’s AI for autonomous weapons and mass surveillance of US citizens. So far, this attempted use of public policy has not resulted in the outcome desired by DoD.
There is also work underway in the US Congress to prevent states from regulating AI. The state of Wisconsin legislature is working on a bill that would limit the liability of AI companies for death and destruction caused by them.
How successful this approach has been outside of the US is hard to evaluate. Both the EU and the UN have created AI study groups. But concrete action by these groups that has a significant effect has not yet been seen.
Other AI companies are calling for the regulation of AI. In the US, they, along with individuals, have created another political action committee to fund politicians running for office who support their position.
Some analysts point to the problems with social media that were exacerbated by a laissez-faire approach. The recent US lawsuit that was successful against Meta and Google tends to support their position.
There is a formal criminal investigation underway in Miami of ChatGPT (OpenAI) being responsible for murder.
There has been talk of adapting to AI job loss by providing everyone a guaranteed minimum income. In the 2020 US presidential primary, there was a candidate who proposed a guaranteed minimum income. His campaign brought a lot of attention to the subject. However, his campaign was not successful, and there has not been that kind of attention since.
The AI 2027 project published a paper in April 2025 on the alignment problem. That is, making sure that AIs incorporate human perspectives and don’t do anything that might cause the end of humanity. The scenario work they published had a significant effect on the AI industry's self-regulation. However, over time, competitive pressures diluted the self-regulation. It also didn’t address the problem of who writes the Spec that AIs are aligned to.
Recent Attempts at Innovations in Public Policy
Glasswing, a project organized by Anthropic to help with society’s adaptation to the power of its new AI, Mythos, is an example of a frontier model developer working to help society adapt to AI. Glasswing brings together Anthropic, large corporations (with emphasis on banks), and cybersecurity defense tool companies whose cybersecurity would be threatened by Mythos to develop adaptations. While these adaptations are being made, Anthropic is holding Mythos off the market. This is an attempt to avoid society having to make any changes by instead finding a technical solution. This is a commendable effort and one that can be a model for others. With public policy encouragement, it could be extended to other areas where AI needs to adapt to society. However, it has very limited applicability to the rest of the areas where society needs to adapt to AI.
