White House and Congress AI Strategy

The president would like to help his campaign donors and his kids by killing off any limits on how AI is used. Since the White House has not espoused a strategy as of yet on placing any guardrails on AI (other than we don’t want any guardrails on AI), the states have stepped in and a few have created laws placing limited guardrails on a few of the very biggest AI companies.

But even those limited guardrails on the very largest companies is making the president unhappy, so he is continuing to try and kill existing and new state laws.

First he tried to get Congress to add in a bill to preempt state laws on AI in the big tax and spend bill passed a couple of months ago. That didn’t work because to many Congress people realized that was not popular with constituents.

Then he tried to get Congress to sneak it in the “must pass” National Defense Authorization Act or NDAA that is almost done and may be voted on this week. Apparently, those same Congress people removed it from there.

The current plan is that he is going to sign an executive order preempting state AI laws.

Which he can do, but an EO can’t preempt a state law.

But what an EO can do is ask the people who work for him in the executive branch to harass states that do pass such laws. One way the EO can do that is, for example, to tell the Justice Department to sue states that pass laws he doesn’t like.

Please remember that the Republicans are supposedly the party that thinks states’ rights is a constitutional mandate. Except when it is inconvenient.

Also remember that if this EO is actually effective, which is unclear, that means that, for example, chatbots having sexually explicit conversations with your young children will be legal again.

Also legal will be AI that encourages your kids to harm themselves, even kill themselves or encourage them to create deepfake porn of other kids and AI algorithms that make decisions about whether you get or keep your job and whether your insurance company will pay your claim. We are already seeing insurance companies using AI to review (and deny) thousands of claims at a rate that exceeds one denial per second.

Obviously, the big tech companies who are creating these AI products would be thrilled if they don’t have to worry about all these pesky matters.

Companies that work to create AI tools responsibly, under this future EO would be at a disadvantage.

The president has also suggested that the EO might withhold unrelated federal funding to states that have AI laws he doesn’t like.

In fairness, this will be other winners if this EO happens and that is outside law firms who represent the states who sue the federal government. That is a very expensive proposition and everyone involved, including expert witnesses, stand to make a lot of money, so, I guess, that is good for the economy. Credit: Route Fifty

But on the other side, lawmakers are asking the White House what their strategy is for countering AI-fueled attacks. They are worried about the emerging threat to the US, both the private sector and public sector, posed by foreign adversaries deploying AI against us. This concern comes from both sides of the aisle. So far, apparently, the White House doesn’t have a plan to deal with this. We have seen this problem already become a reality and it will only get more effective. But, if the White House strategy is no strategy, the Chinese and North Koreans say thank you. Credit: Cyber Security Dive

By the way, I have no objection to the feds actually implementing an AI regulatory strategy that works, but I think that is only a dream.

by