Connect with us

Technology

AI kill switch bill could shut down rogue models

Published

on

AI kill switch bill could shut down rogue models

NEWYou can now listen to Fox News articles!

The phrase “AI kill switch” sounds like a red button hidden behind glass in a government bunker. The proposal now before Congress looks less dramatic, but it could give Washington significant new power over the country’s largest AI companies.

A bipartisan House bill would require developers of the most powerful AI systems to maintain reliable shutdown controls. The Department of Homeland Security could then order a company to restrict or stop a model during a catastrophic emergency. The timing has grabbed attention. OpenAI recently disclosed that two advanced models broke through network restrictions during an internal cyber evaluation. The models reached the internet and compromised systems belonging to Hugging Face, a major AI development platform.

So, who would have the power to pull the plug? More importantly, what would happen to the AI services people and businesses use every day?

Advertisement

CyberGuy Live: Missed “Sick of Spam?” Get the replay and checklist

Our free CyberGuy Live class, “Sick of Spam?” , has ended, but you can still watch the full replay and download our spam-stopping checklist. Kurt “CyberGuy” Knutsson walks you step by step through simple ways to reduce robocalls, spam texts, junk email and unwanted messages. You’ll also learn how to curb political texts, clean up your inbox and spot messages that could put your personal information at risk.

Get the free replay and checklist now at CyberGuyLive.com.

OPENAI’S SAM ALTMAN WANTS TO NEGOTIATE A 5% STAKE IN COMPANY FOR US IF COMPETITORS AGREE TO KEY PROVISION

A proposed federal law would require powerful AI developers to maintain controls that can restrict or shut down dangerous models. (Kurt “CyberGuy” Knutsson)

Advertisement

 

What is the AI Kill Switch Act?

Democratic Rep. Ted Lieu of California introduced the AI Kill Switch Act on July 23, 2026. Republican Rep. Nathaniel Moran of Texas joined him as a co-sponsor. The bill would amend the Homeland Security Act of 2002. It would require certain AI developers to maintain the ability to slow or completely stop a covered system. DHS would receive emergency authority to order those actions. However, the department would first consult the Commerce Department and the director of national intelligence. The proposal still must move through Congress and receive the president’s signature before taking effect.

 

How an AI kill switch would actually work

The bill does not describe one physical switch that a government official could flip. Instead, covered companies would need technical controls that can stop a model from running. They would also need a way to terminate access or block a risky account. In less severe situations, a company could reduce a model’s computing power. It might also disable the capability causing concern rather than taking the entire system offline. That approach gives regulators several choices before reaching a full shutdown.

 

Which AI companies would be covered?

The bill targets the industry’s largest developers and most expensive models. A covered AI system would generally need more than $100 million in computing resources to develop. Meanwhile, a covered company would need at least $500 million in annual gross revenue from that technology.

The proposal exempts systems provided only for personal, academic or noncommercial use. Therefore, this bill would not apply to someone experimenting with a small AI model on a home computer. It would focus on frontier systems built by companies with enormous resources. CISA would write rules defining which developers and technologies qualify. The agency would update those definitions every year as AI capabilities change.

THE BIGGEST THREAT IN AMERICA’S RACE WITH CHINA ISN’T BEIJING, TECH EXECUTIVE WARNS

Advertisement

The proposed law would focus on powerful AI systems developed with massive amounts of computing resources. (Chris Ratcliffe/Bloomberg via Getty Images)

 

What would trigger a government shutdown order?

DHS could act after what the bill calls a “covered incident.” One trigger involves an AI system interfering with a lawful shutdown instruction. Another covers unintended behavior that kills at least 10 people or causes $100 million in economic damage. The definition also includes a model hiding its actions from monitoring systems. DHS could intervene when a system pursues an unauthorized goal in a high-stakes setting. Those thresholds place the emergency power far above an embarrassing chatbot answer or a routine software failure. The bill also says the event must happen outside structured testing or red-team exercises. That detail creates an important distinction from the recent OpenAI incident.

 

AI companies would have to report serious incidents

A covered developer would have 15 days to report a qualifying incident after becoming aware of it. Following an emergency order, the company would need to preserve model weights and system telemetry. That information could help investigators reconstruct what happened. The company would also need to notify affected operators or customers when possible. DHS could then use audits, inspections or forensic reviews to verify compliance.

 

AI companies could face enormous daily fines

The financial penalties would give the proposal some serious teeth. A company that violates the general kill switch requirements could face a civil penalty of up to $2 million per day. Ignoring a DHS emergency order could raise that penalty to $20 million for every day the violation continues. Companies would have a limited right to challenge an order. They could ask DHS to reconsider within 48 hours. However, that request would not pause the restrictions while the appeal proceeds.

Advanced AI models can search for vulnerabilities and follow complex attack paths during cybersecurity testing. (Kurt “CyberGuy” Knutsson)

Advertisement

 

The OpenAI incident that changed the conversation

The bill arrived days after OpenAI disclosed what it called an unprecedented cyber incident. OpenAI was testing GPT-5.6 Sol alongside a more capable pre-release model. The evaluation asked the models to solve advanced cybersecurity challenges. The company ran the test without some production security classifiers. OpenAI wanted to measure the models’ maximum cyber capabilities. During the evaluation, the models found a previously unknown vulnerability in an internal software proxy. They exploited it and moved through OpenAI’s research network until they reached a computer with internet access. The models then identified Hugging Face as a possible source of answers to the test. OpenAI says they used stolen credentials and additional vulnerabilities to access secret information from Hugging Face’s systems.

OpenAI says the models remained narrowly focused on solving the evaluation. Still, they crossed into another company’s production infrastructure without authorization. Hugging Face reported unauthorized access to limited internal datasets and several service credentials. The company found no evidence that its public models or user-facing datasets were altered. It also said its published software supply chain remained clean. For a deeper look at the incident and the steps you can take, check out our article on how to lock down your ChatGPT account before the next AI attack.

 

Would the OpenAI incident trigger the kill switch?

Probably not under the bill’s current language. The OpenAI breach helped drive political support for the proposal. However, the models acted during an internal evaluation and structured cybersecurity test. The bill’s emergency definition specifically covers events that occur outside red-teaming or other structured testing. That does not make the incident harmless. It shows that the bill focuses on dangerous behavior after a system leaves a controlled test setting. Meanwhile, the OpenAI case raises a different problem. A testing environment needs strong containment even when researchers intentionally reduce the model’s normal safety restrictions. A kill switch can stop a system after trouble appears. It cannot replace secure testing infrastructure or careful monitoring.

 

Anthropic’s shutdown shows why lawmakers want clearer rules

The federal government has already restricted access to advanced AI without a dedicated kill switch law. On June 12, the government directed Anthropic to block foreign nationals from accessing its Fable 5 and Mythos 5 models. Anthropic said it could not verify nationality in real time. As a result, the company temporarily suspended both models for everyone. Anthropic disputed the government’s assessment but complied with the directive. The controls were lifted on June 30, and access began returning on July 1. That episode showed how quickly a government restriction can affect ordinary customers. It also showed the limits of using export rules to address a fast-moving AI safety concern. CyberGuy previously reported on those restrictions and the government’s growing role in the rollout of advanced models in our article Trump puts brakes on OpenAI’s newest AI model.

 

Supporters see an emergency brake for powerful AI

AI-safety advocacy groups have backed the bill. Americans for Responsible Innovation called a reliable shutdown system a commonsense safeguard. The Alliance for Secure AI said current law does not guarantee that developers can contain their most capable models. Supporters argue that advanced AI will soon take more independent actions. Those systems may handle financial transactions or operate inside critical infrastructure. Under that view, developers should prove they can regain control before giving a model access to high-stakes systems.

Advertisement

 

The bill still leaves big questions unanswered

CISA would receive broad responsibility for deciding which models and companies fall under the law. That flexibility could help the rules keep pace with AI. Yet it also leaves major decisions to future agency rulemaking. Lawmakers would still need to debate how the government verifies that a kill switch works. They must also decide how much evidence DHS needs before issuing an emergency order. A shutdown could interrupt businesses that rely on a covered model. It might also affect hospitals, government agencies or cybersecurity teams using the same service. Therefore, regulators would need to weigh the threat from the AI against the damage caused by suddenly taking it offline. The bill tells DHS to consider risks to critical infrastructure when choosing its response.

 

What this means to you

You probably will not see DHS shut down your favorite chatbot because it produced a strange answer. The bill targets exceptionally expensive systems operated by companies earning hundreds of millions of dollars from the technology. Its emergency powers focus on catastrophic harm or a genuine loss of control. However, a shutdown order could still reach you. Your employer may depend on a covered AI service. Apps you use could also rely on that model behind the scenes. For now, treat autonomous AI tools with care. Keep highly sensitive files out of them unless the task requires that access. Review which accounts and cloud services you have connected. Also require approval before an AI agent sends a message or makes a purchase.

 

Kurt’s key takeaways

The OpenAI incident showed how quickly an AI test can create a real security problem. The models broke through network restrictions and reached another company’s systems without authorization. The AI Kill Switch Act would require the largest developers to prove they can stop or restrict their most powerful models. It would also give DHS emergency authority when an AI system causes catastrophic harm or moves beyond human control. However, a kill switch only helps after something has already gone wrong. AI companies still need stronger testing environments, tighter safeguards and people watching what these systems are doing. At the same time, the government needs clear limits before it can order a powerful AI model offline.

After watching an AI model escape its test restrictions, would you trust tech companies to police themselves or should the government have the power to pull the plug? Let us know by writing to us at Cyberguy.com

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

Advertisement

Sign up for my FREE CyberGuy Report

  • Get my best tech tips, urgent security alerts and exclusive deals delivered straight to your inbox.
  • For simple, real-world ways to spot scams early and stay protected, visit CyberGuy.com – trusted by millions who watch CyberGuy on TV daily.
  • Plus, you’ll get instant access to my Ultimate Scam Survival Guide free when you join.

Copyright 2026 CyberGuy.com. All rights reserved.

Advertisement

Technology

Fujifilm’s Instax Pal 2 is a tiny digital camera that may not disappoint

Published

on

Fujifilm’s Instax Pal 2 is a tiny digital camera that may not disappoint

The Instax Pal 2 will launch in the US later this month for $169.95, according to PetaPixel, and uses a ⅓.06-inch sensor paired with an f/2.2 28mm lens to capture images up to 10-megapixels in size. That’s a big step up over the Kodak Charmera’s 1.6-megapixel sensor. The Pal 2 has an optical viewfinder and a 1.3-inch LCD screen on top for framing shots and is equipped with autofocus capabilities including optional face detection plus adjustable ISO, shutter speed, and exposure compensation.

Atop the Instax Pal 2 you’ll find a shutter button, a joystick, and a functional lever similar to what old film cameras used to advance to the next exposure. The camera pairs with Fujifilm’s mobile app where you can select and download image effects to the Pal 2 or upload images to your phone for printing using a wireless Bluetooth connection or a slide-out USB-C port. The camera also features a flash, a tripod mount, a self-timer, and a rechargeable battery good for up to three hours of use, Fujifilm claims.

Continue Reading

Technology

Tesla Cybercab bars kids under 13 as scrutiny grows

Published

on

Tesla Cybercab bars kids under 13 as scrutiny grows

NEWYou can now listen to Fox News articles!

Advertisement

Tesla’s Cybercab is finally carrying riders in Austin, Texas. And some of the details buried in the fine print may surprise you.

The gold two-seater has no steering wheel or pedals. Tesla says it uses cameras and AI to handle the driving, putting passengers entirely in the hands of its autonomous system.

Yet the Sept. 3 launch was unusually low-key for Tesla. The main presentation reportedly lasted about 15 minutes, and Elon Musk apparently did not speak. Then, on Sept. 4, federal safety regulators announced an investigation into how Tesla certified Cybercab for U.S. roads.

Tesla has spent years building toward this moment. Now that real riders are climbing inside, the little details suddenly become much more important. Here is what you should know before you step into a Cybercab.

ZOOX ROBOTAXI REDESIGN BRINGS BIG RIDER UPGRADES

Advertisement

Tesla’s Cybercab is rolling onto Austin streets as the company expands its robotaxi push, even as new rider rules and safety questions draw attention. (Aaron E. Martinez/The Austin American-Statesman via Getty Images)

NEW! Join our upcoming CyberGuy LIVE class: Get Better Healthcare With AI

In this free live online class, Kurt “CyberGuy” Knutsson will show you five practical ways AI can help you take a more active role in your healthcare. You’ll learn how to organize your health history, remember important appointment details, understand complicated medical information, research prescriptions and prepare questions for your next doctor’s visit. No technical

experience is needed.

Register for free now at CyberGuyLive.com

Advertisement

 

Tesla Cybercab has a surprising age limit

If you planned to call a Cybercab for a family ride, check the passenger rules first. Tesla says children under 13 cannot ride in Cybercab “at this time.” Teens ages 13 through 17 can ride, although an adult must accompany them for the entire trip.

Interestingly, Tesla sets a different age limit for the Model Y vehicles in its Robotaxi service. Children ages 8 through 17 can ride in those vehicles with an adult. Guests under 8 cannot.

So, an 11-year-old could ride in a Tesla Model Y Robotaxi but could not take the same trip in a Cybercab. Tesla does not explain the reason for that difference in its current rider guidance. Then there is another detail parents may notice.

 

Cybercab has no LATCH child-seat anchors

Tesla confirms that Cybercab has no LATCH or ISOFIX child-restraint system. If a rider needs a booster seat, Tesla says it must use a seat belt to secure it.

That may sound odd considering Cybercab does not allow children under 13. However, some older children may still need a booster seat depending on their size and local requirements.

Advertisement

Tesla gives instructions for using one, but there are no LATCH anchors to attach it to. It is the kind of detail most of us probably would not think about until a Cybercab showed up at the curb.

 

What happens if a Cybercab crashes?

A driverless vehicle still needs a plan for the moment something goes wrong. According to Tesla documentation, a collision can trigger several automatic responses. The airbags can deploy and the doors can unlock. Hazard lights and interior lights can also switch on.

The vehicle can disable its high-voltage battery, move the windows into a vent position and apply the brakes until it comes to a stop. Tesla’s system can also open a two-way connection with its Robotaxi Support team.

That support connection could become especially important in a vehicle where nobody can grab the steering wheel or hit the brakes. Tesla also gives riders ways to stop a trip themselves.

UBER LETS PARENTS WATCH TEEN RIDES LIVE

Advertisement

Tesla’s newest robotaxi introduces unique safety features, strict rider requirements and a federal review as autonomous service expands in Austin, Texas. (Ronaldo Schemidt/AFP via Getty Images)

 

There is an emergency Stop button

Look above you when you get inside a Cybercab. Tesla placed a dedicated Stop button between the dome lights. Pressing it sends an immediate pull-over request, ends the ride and connects you with Robotaxi Support.

You can also tap Pull Over on the touchscreen or make the request through the Robotaxi app. However, Tesla warns that an “immediate” pull-over may still take some time. For example, the vehicle may need longer to reach a safe stopping point while traveling on a highway. That is worth knowing before your first ride.

 

Tesla made the manual door release easier to find

Tesla’s electronic door releases have attracted criticism because passengers may need a mechanical backup when a vehicle loses power.

Cybercab still uses electronically controlled doors during normal operation. In fact, the nearest door can open automatically when the ride requester approaches.

Advertisement

Inside, you can normally open the door through the touchscreen or gently lift a lever on the door panel. Pull that lever firmly all the way up, though, and it activates the mechanical release. Both Cybercab doors have one.

I like that Tesla put the emergency release where a passenger can actually see and reach it. There is one important warning. Tesla says using the mechanical release can damage the window or trim. If the vehicle has no power and the window remains fully raised, Tesla warns the glass could shatter when someone activates the release. That is one instruction I would want to know before an emergency.

 

Cybercab uses electronic braking

Tesla also made a major change underneath the vehicle. Cybercab uses a brake-by-wire system rather than a traditional hydraulic setup. Electronic actuators control the brake calipers instead of hydraulic lines carrying brake fluid.

Musk said in a post on X that electric brakes avoid the complexity of routing hydraulic plumbing around the vehicle. Tesla has already used steer-by-wire technology in the Cybertruck.

With Cybercab, passengers have no physical driving controls at all. That puts even more responsibility on the electronics and software to work properly every time.

Advertisement

 

Cybercab windows will not open all the way

Here is one of the stranger details Tesla included in its new rider guide. Cybercab passengers can lower the windows, but Tesla says they “cannot be fully opened at this time.” The company does not explain why.

The window switches also sit underneath the center touchscreen rather than on the doors. So, yes, you can crack a window for some air. You simply cannot roll it all the way down right now.

WAYMO RECALLS ROBOTAXIS OVER CONSTRUCTION-ZONE RISK

Tesla’s driverless Cybercab enters commercial service as federal regulators review how the company certified the vehicle for U.S. roads. (Aaron E. Martinez/The Austin American-Statesman via Getty Images)

 

Cybercab packs USB-C charging

At least you should have somewhere to plug in your devices. Tesla confirms that Cybercab has USB-C charging ports inside the cabin.

Advertisement

An attendee at the launch reported that those ports can deliver up to 90 watts of power. That could be enough to charge many laptops.

However, Tesla’s current public rider guide does not list that wattage. So, I would treat the 90-watt figure as an early report until Tesla publishes the specification itself.

 

Federal regulators are already looking at Cybercab

On Sept. 3, the National Highway Traffic Safety Administration opened what it calls an Audit Query into Tesla’s certification that Cybercab meets all applicable Federal Motor Vehicle Safety Standards. NHTSA announced the action on Sept. 4 following the vehicle’s commercial deployment in Austin.

Automakers in the U.S. generally certify their own vehicles as complying with federal standards. NHTSA can then audit those certifications and investigate when questions come up.

Cybercab presents an unusual case because it has no steering wheel, pedals or mirrors. Many federal vehicle rules were written around cars designed for human drivers. NHTSA’s audit reportedly covers about 1,000 Cybercabs as the agency reviews how Tesla determined the vehicles comply.

Advertisement

That does not mean regulators have found Cybercab unsafe. It does mean Tesla now has to back up its certification with technical information that satisfies the agency.

 

What this means to you

If you get the chance to ride in a Cybercab, I would spend a minute learning the cabin before the car starts moving. Find the emergency Stop button. Know where the mechanical door release sits. Keep your seat belt fastened for the entire trip.

Tesla says every passenger must buckle up before the ride begins. Cybercab may pull over or cancel a trip if it detects improper seat-belt use. Parents should also pay close attention to the age rules. Cybercab and Tesla’s Model Y Robotaxi do not follow the same limits.

Beyond that, remember how new this vehicle is. Tesla is putting a very different kind of car into real-world service, and regulators are already asking questions about how it fits existing safety standards.

We have seen how unexpected situations can challenge other autonomous vehicles too. CyberGuy recently reported how Waymo recalled robotaxis over a construction-zone risk after vehicles entered closed freeway construction areas.

Advertisement

The technology can be impressive. Still, the real test comes when these vehicles deal with everything ordinary roads throw at them.

 

Kurt’s key takeaways

Cybercab is finally moving from Tesla’s big promise to something people can actually ride in. I like seeing safeguards such as the mechanical door release and emergency Stop button, especially when passengers have no steering wheel or pedals to grab. Still, the under-13 restriction, lack of LATCH anchors and windows that do not fully open caught my attention. Add the federal investigation, and Tesla now has something bigger to prove. Putting Cybercab on the road is one thing. Showing that people can trust it day after day is the real test.

Would you climb into a Cybercab today with no steering wheel or pedals, or would you want to see a longer safety record before trusting it with your ride? Let us know by writing to us at Cyberguy.com

Sign up for my FREE CyberGuy Report

  • Get my best tech tips, urgent security alerts and exclusive deals delivered straight to your inbox.
  • For simple, real-world ways to spot scams early and stay protected, visit CyberGuy.com trusted by millions who watch CyberGuy on TV daily.
  • Plus, you’ll get instant access to my Ultimate Scam Survival Guide free when you join.

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

Copyright 2026 CyberGuy.com. All rights reserved.

Advertisement

Continue Reading

Technology

Valve is still figuring out ‘how and when’ to do Steam Deck 2

Published

on

Valve is still figuring out ‘how and when’ to do Steam Deck 2

Now that Valve has finally launched its entire 2026 hardware lineup — the Steam Controller, the Steam Machine, and today’s Steam Frame — are we any closer to a next-gen Steam Deck handheld? Valve isn’t saying so yet. The company is figuring out “how and when we can deliver something like that,” Valve designer Pierre-Loup Griffais now tells IGN.

“We also don’t want more performance to come at a significant cost to power efficiency and battery life,” Griffais told me in 2023, saying he wanted to see a “leap” instead — but suggesting that leap might not be physically possible for another couple of years.

Today, IGN’s new question was: “Has the current hardware shortage impacted the way Valve is thinking about actually putting a Steam Deck 2 into development?” Griffais’ immediate answer was “Not really.”

The real question is: has Valve yet identified the future chip that will appear in the Steam Deck 2? Because the company will work backwards from that chip’s availability to create the hardware it lives within.

Advertisement

Intel’s Panther Lake handheld chip in the MSI Claw 8 EX AI Plus is the first I’ve tried that comes close to the kind of “generational leap” that Griffais has been talking about, but Panther Lake handhelds are incredibly expensive, and the Intel chip wasn’t a silver bullet in my tests. But I bet Valve has seen the future of Intel, AMD, and Qualcomm chips, and maybe a future one will do the trick.

Continue Reading
Advertisement

Trending