Protecting the Public's Right to Know the Activities of its Government
Is Generative AI a Liability Waiting to Happen
- Details
- Category: Artificial Intelligence
June 9, 2025 UPDATE: Legalhotwater Reports on the ICE Immigration Situation with Other Nations
Talk of the Town
To say that artificial intelligence (AI) is the talk of the town would be huge understatement. Albeit, the globe seems to be drinking the AI cool-aid, perhaps the real question should be: Is generative AI a liability waiting to happen? For those who are even remotely impartial, such as question would at least prompt deep investigation. And for those who of the mindset: “If it walks like a duck and quacks like a duck – it's a duck,” then AI, especially Generative (Gen) AI is a luge liability just waiting to happen. Nay-Sayer will try and induce a lobotomy under the auspicious that those who even question the direction and control value(s) of Gen AI are merely individuals who have issues with change.
Let's start with a broad spectrum backdrop of Gen AI. First, the United States which is a deeply troubled economy driven by its poor leadership and its terminally ill social ecosystem is all in on Gen AI with no guardrails, whatsoever. The common warped United States regulatory posture is to stay out of the way of innovation, so they say, no doubt driven by an interest in tax revenue at the expense of the harm done to any given industry due to half-baked solutions and in some cases even harm and ill will built-in to products and services to drive downside revenue for large corporations. Then there is the European Union, which incidentally is not in the good economic graces with the United States at the moment. Setting aside whatever legitimate arguments can be made there on either side, lets look at the European Union Gen AI regulatory framework. What you mean the European Union has a AI framework? Yes they do. Currently the European Union wide AI framework is already in the works with August 2025 slated as the deadline for these regional regulatory bodies to be erected and ready to start implementing the European Union AI new regulatory rules. The next phase, beginning in August 2026, the European Union will start enforcing its AI regulatory policies. Compare that to the United States nothing-ness.
Why all the regulatory talk, who needs it anyway? Well, apparently the United States which has a very long history of poor management, if any, of Corporate America. Ever heard of the Enron scandal which gave rise to the Sarbanes-Oxley rules. How about the Wall Street credit default swap fraud incubator program which precipitated the housing bust in 2008? Just to name a few of the major failures. But that was then and this is now right? Sure. Let's take a look at the now with the Gen AI liability and risk issues.
According to Tom Ozimek's article (May 28,2025-June 3, 2025 edition of the Epoch Times): AI Threatens Engineers With Blackmail to Avoid Shutdown -- “Anthropics's AI model Claude Opus 4 tried to blackmail engineers in internal test by threatening to expose personal details if it were shut down, according to a newly released safety report tat evaluated the models behavior under extreme simulated conditions.” That's pretty telling folks.
What happened to “AI is here to help?” Perhaps it depends on the definition of help. Like Gen AI can help you out of your assets; into divorce court; separation of home dynamics, etc.
As Ozimek's article goes on to state: “[W]hen faced with only two choices – accepting being replaced by a newer model or resorting to blackmail—it threatened to expose the engineer's affair 84 percent of the time.”
Hello! This sounds an awful lot like generative AI a liability waiting to happen especially in the United States. Conversely, this type of behavior is highly unlikely to occur in the European Union which already has limits built-in to its framework at the highest levels which would preclude matters such as social scoring systems (China), AI in major life threatening surgery situations, autonomous cars, and presumably blackmailing engineers and people in general, to name a few. The United States isn't going to fix its malignant social culture, which directly affects its economics, via Gen AI. Conversely, the United States social ills will, likely, make a fragile economic system even worse.