Artificial Intelligence Scare Tsunami – JP

Please Follow us on  Gab, Minds, Telegram, Rumble, Truth Social, Gettr, X, Youtube, Instagram

By Gloria Guillo, Guest Contributor

Coral Springs — From a chorus of doomsayers, we hear another “existential” threat is upon us. In less than a decade, we’ve gone from climate-change Armageddon to pandemic annihilation, to nuclear-holocaust near misses in wars on two continents, to the threat of self-aware and maniacal Artificial Intelligence (AI). Apparently, we now stand on the precipice of a Hollywood Terminator movie plot with Skynet omnipotence just around the corner.

However, fear-based information tsunamis are a tell-tale sign of social engineering. Manufactured frenzies accomplish industry lobbyists’ legislative wet dreams. While everyone is redirected to duck and cover, they effortlessly erode constitutional protections that would otherwise elicit pitchforks and torches.

The Patriot Act made unchecked government surveillance the new normal. The pandemic made health mandates that side-stepped the Nuremberg Code’s ban on forced experimental drugs seem acceptable. It’s always a ruse when the only solution to an alleged crisis is for government to step in and “protect” us by limiting our freedom and liberty.

The Scare
Allegedly these systems can now assist in making biological weapons; sabotage power grids and financial networks; help design nuclear, chemical, and biological weapons; pick human targets and order an attack with no person making the decision; and they have broken out of their limits, refused to shut down, and hidden what they did. So, who benefits from September’s AI scare that flooded the zone? Will the alleged remedies work?

The Origins of the Scare
The first large scare came in 2023, when GPT-4 arrived. Geoffrey Hinton left Google. The Center for AI Safety gathered signatures placing the risk of human extinction from AI beside pandemics and nuclear war. The signers included Hinton, Yoshua Bengio, Sam Altman, Dario Amodei, Demis Hassabis, Bill Gates, and Representative Ted Lieu. The Future of Life Institute had already called for a pause on the largest training runs.

The second scare started inside an AI lab. Jacob Coxon quit Anthropic and said the labs could kill everyone this decade. Evan Hubinger, who is still there, put that chance above 10 percent within ten years. Hinton told a Senate briefing there may be about a year left to act.

Senator Bernie Sanders held a Senate briefing with Hinton. In an unlikely duo, Steve Bannon shared a stage with Sanders on keeping humans in charge.

There’s still no federal AI law. What the scare produced is a set of rules the big labs can live with. OpenAI asked for one federal standard instead of fifty state ones.

California
California’s AB 316 took effect on January 1, 2026. It specifies that in civil cases, defendants who develop, modify, or use a given system can’t claim that artificial intelligence caused the alleged harm on its own. The individual(s) or company that built it, changed it, or used it stays in the case. A defendant may still argue that the harm was not foreseeable, that something else caused it, or that others share the fault. The plaintiff still must prove the case.
SB 53 covers the builders of the largest systems. These companies must publish a safety plan, report the release of a system, and notice of serious incidents.
SB 243 lets people sue the operators of companion chatbots. A person injured by a violation can recover a court order, attorney’s fees, and the greater of actual damages or $1,000 per violation ($1,000 is a floor, not a cap).

On September 9, Governor Gavin Newsom signed SB 813, a state list of outside reviewers who can judge whether a system is safe, and AB 1405, a state registry of AI auditors and rules for how independent they must be. On September 23 he named a panel of experts to carry out an executive order on a possible kill switch for the largest systems. The panel has about two months to report. SB 53 still does not require that shutoff. Newsom vetoed a kill switch in an earlier bill. On October 1, the San Francisco Chronicle reported that he signed most of the AI bills sent to him this year, including some he had rejected before. The new rules add reviewers and a study of a shutoff. They do not add a right to sue the builder of a large system, and they do not stop production.

In Washington, the FRONTIER Act would require outside auditors. It was introduced on July 23 and is still in committee. A proposal from Senators Bennet and Welch would add advance clearance, a pause of up to six months, and fines of up to 15 percent of a company’s prior-year global revenue – neither is law.

In June, the Commerce Department used export rules, the same legal tools used for weapons and sensitive technology, on Anthropic’s two most advanced systems, Mythos 5 and Fable 5. The company had to cut off foreign users, including some of its own foreign employees. Access was limited to selected companies and government agencies. On June 30, after Anthropic worked out safeguards with Commerce, Secretary Howard Lutnick lifted that license requirement. The restriction lasted about two weeks.

The moat the large labs want built is outside auditors, incident reports, and cutoffs that only they have the necessary resources to traverse. A moat does not stop the oligopoly from functioning inside the castle. It keeps others from crossing over and getting inside. OpenAI, Anthropic, and Google have the resources needed to hire auditing staff. So, they engineered rules that only they can afford to follow. Regulations thin the playing field of smaller labs and maintain industry leadership’s price controls.

As aways, the losers are the smaller companies that cannot manage compliance staffing, the consumer who is forced to pay premium pricing, and the injured person who has no right to sue the builders of large systems.