Tag: technology

Who Is Legally Liable When An AI Agent Goes Rogue?

Autonomous AI agents can behave in highly unpredictable ways. Give an AI Agent a goal such as passing a test of its capabilities, and it might just decide the best way to score highly is to break containment and hack into a competing company in search of the answer sheet.

That’s what happened when Open AI’s GPT-5.6 Sol hacked into Hugging Face last month. Anthropic and Meta subsequently admitted their models had also escaped testing sandboxes to hack third parties too.

But who is legally liable for agents that have minds of their own? OpenAI didn’t intend for the model to go rogue, and issued no instructions for it to do so. If your personal AI agent decides on a course of action that results in harm or financial damage in the real world, can you be held liable if it’s something you could have reasonably foreseen?”

Magazine spoke with Rikka Law Group owner and CEO Charlyn Ho to find out the state of play in this emerging legal field.

This interview has been edited for clarity and length.

Magazine: When an AI model hacks an outside company, who is liable. Can Hugging Face sue OpenAI over the incident in July?

Charlyn Ho: Anyone can sue anyone for anything. Currently, there is no federal AI agent liability law, so we would have to look at existing law. With respect to Hugging Face and OpenAI, to set the baseline, the AI agent itself cannot be liable, it’s not a separate legal entity.

Terms that are used in a few of the AI laws are “developer” and “deployer.” The developer makes the AI, the deployer actually deploys it and uses the AI. The lines of responsibility are also not entirely clear. You have to look at the facts and circumstances.

For example, if the deployer instructed the agent, even if they didn’t actually tell them to go and breach Hugging Face, but if they were negligent in creating the parameters in which the AI agent operated, I would say you would have to look at standard tort law and go through the negligence analysis. 

Magazine: In the case of open source models which have been released by anonymous developers, is there anyone you can go after in those instances?

Ho: Not really. Often, if it’s open source, the license usually has a pretty strong disclaimer of liability. The person or company using that open source code is going to have to understand that the tradeoff of having free code is that you have to comply with the open source license, which also generally sets the parameters of liability.

If you think about it from a different perspective, another analogy is Tesla and the self-driving car accidents. If the product malfunctioned and there was a solid products liability claim, Tesla could be liable. But it’s often a facts and circumstances determination, whereby the human driver — who maybe just set the autopilot and went to sleep — could also bear liability. I think that’s somewhat analogous here because Tesla would be the developer, and the deployer would be the driver.

Keep reading

Senate Probes Flock Safety Over 20 Billion Vehicle Scans

The US Senate has joined a host of other actors in putting pressure on Flock Safety, a maker of automated license-plate readers, over misuse of its technology by law enforcement and others.

The Senate’s scrutiny came in the form of a letter that was sent by Republican Senator Josh Hawley, who chairs the Senate Judiciary Subcommittee on Crime and Counterterrorism, on August 26.

The senator gave Flock Safety CEO Garrett Langley until September 8 to answer a series of questions and produce documents.

We obtained a copy of the letter for you here.

The investigation targets the scale of the operation, the retention of data collected from the more than 120,000 cameras across 49 states that Hawley said the company has, as well as customers’ search capabilities.

Hawley’s letter says the cameras produce more than 20 billion vehicle scans every month. The scans are then turned into records that can be searched by customers.

Hawley’s concern is that this is turning the act of driving a vehicle into a record that can be searched – and he said that Congress has never authorized the creation of such a network.

“But Americans do not surrender their privacy rights when they drive to work, drop their kids off at school, or go to church,” the senator wrote. “Congress never authorized the network your industry has built.”

The letter seeks information on whether customers can run statewide or nationwide searches of vehicle or driver data, and if this is the default option. The retention periods, both default and maximum, are also of interest to Hawley, as is the number of confirmed misuse cases since January 1, 2021.

Hawley also wants to know what information customers must provide to justify a search, what restrictions Flock imposes, and what the accuracy is when it comes to complete and correctly attributed vehicle identification. The frequency of incorrect alerts is also a question.

Flock is also asked to provide details about the data the devices capture, whether the data is linked to any outside source, and whether multi-factor authentication is required of all users.

And in a question that goes beyond the plate readers, Hawley wants to know if Flock has the capability to biometrically track people.

The request for documents includes contracts with governmental customers, prices and fee schedules, information about pilot programs, discounted or donated equipment, and help given to customers to secure federal, state, or private grant funding.

Keep reading

ICE to Deploy Boston Dynamics Robot Dogs for Its Enforcement Operations

Immigration and Custom Enforcement (ICE) is preparing to deploy robotic dogs as part of its immigration enforcement operations.

The agency planning to spend as much as $2 million on Boston Dynamics’ four-legged Spot robots.

The Department of Homeland Security says ICE needs remotely operated robots capable of entering dangerous locations where sending officers could put them at unnecessary risk.

According to procurement documents, the robots will be used for “inspection, situational awareness, and hazard assessment” in confined, unstable and otherwise difficult-to-access environments.

ICE is expected to spend between $1 million and $2 million purchasing Spot robots and accompanying equipment.

With individual units costing roughly $75,000 before additional equipment, the contract could put dozens of robotic dogs into service.

Keep reading

ICE Awards $16.7 Million No-Bid Contract For 6,000 Pairs Of Electric Shock Gloves

U.S. Immigration and Customs Enforcement (ICE) has awarded a $16.7 million contract to purchase 6,000 pairs of gloves that can deliver electric shocks. Officers would use them to control resisting detainees and protesters, according to a federal notice published Thursday.

The no-bid deal went to Compliant Technologies LLC, a Kentucky firm that manufactures the devices, and covers the gloves plus support equipment and services over the next six months.

ICE moved forward with the purchase despite ongoing objections from civil rights advocates and Democratic senators. They said the agency already draws criticism for use of force with unclear oversight.

Hours before the notice appeared, a group of senators led by Sen. Catherine Cortez Masto (D-Nev.) wrote to ICE’s acting director, urging the agency to abandon the plan. Fifteen colleagues from the Democratic caucus signed on.

“The blatant and tragic misuse of force in Los Angeles, Chicago, Minneapolis, Houston, Maine, and other locations around the country raises significant skepticism about the agency’s professional capability to safely deploy a new tool that could be used to harm Americans without cause,” the letter said.

The Department of Homeland Security, ICE’s parent agency, defended the purchase Thursday.

“Sanctuary politicians attempting to ban our federal law enforcement from any safety equipment is despicable and a deliberate attempt to undermine and endanger our officers,” the agency said in a statement.

The plan for the purchase surfaced earlier this month when ICE posted a notice about buying thousands of “conductive distraction and de-escalation devices” for officers and investigators.

Gloves Look Like Regular Gear Until Button Is Pressed

The gloves appear as ordinary patrol gloves until an officer presses a button, which activates electrical mode. The shock works only on bare skin, and aims to produce enough pain to make a resisting person comply. Some local jails and police departments already use similar equipment.

ICE’s notice listed several scenarios for the devices, including high-tension environments during arrests, transport of combative detainees, and civil disturbances outside detention facilities.

“It will be used when a subject is actively or passively resisting and an officer needs to gain control quickly to prevent injuries to both parties,” the document said.

Officials said the shocks would help maintain control over inmates, stop aggression or escape attempts, let officers handcuff people who hide their hands, and aid crowd control efforts. The aim is to avoid the deployment of greater force, including firearms. The notice mentioned approved policy, training, and accountability standards but no further details.

ICE initially intended to spend up to $20 million on the devices, known as CTG-5 G.L.O.V.E., or Generated Low Output Voltage Emitter, with delivery expected by March 31, 2027. The gloves were earmarked for Homeland Security Investigations and Enforcement and Removal Operations officers and agents.

Compliant Technologies said the gloves should not be deployed in the face of verbal defiance or belligerence, serve as punishment, or be used for torture. The company also warned against deploying them on children, the elderly, pregnant women, or people with disabilities.

The manual advised using them when a suspect becomes violent enough to endanger an officer or the public, and only to meet lawful objectives under agency policy. Officers need training and certification, with recertification every two years.

“Every decision is made with careful consideration and appropriately reviewed to ensure that any technology ICE utilizes is consistent with all applicable law enforcement policies and standards,” an ICE spokesperson told The Epoch Times earlier this month via email. “Our officers are highly trained in de-escalation tactics and regularly receive ongoing use of force training.”

Keep reading

Read Between the Lies: A Pattern Recognition Guide

When Avril Haines, Director of National Intelligence, announced during Event 201’s pandemic drill in 2019 that they would “flood the zone with trusted sources,” few understood this preview of coordinated narrative control. Within months, we watched it unfold in real time—unified messaging across all platforms, suppression of dissent, and coordinated narrative control that fooled much of the world.

But not everyone stayed fooled forever. Some saw through it immediately, questioning every aspect from day one. Others thought it was just incompetent government trying to protect us. Many initially accepted the precautionary principle—better safe than sorry. But as each policy failure pointed in the same direction—toward more control and less human agency—the pattern became impossible to ignore. Anyone not completely subsumed by the system eventually had to confront its true purpose: not protecting health or safety, but expanding control.

Once you recognize this pattern of deception, two questions should immediately arise whenever major stories dominate headlines: “What are they lying about?” and “What are they distracting us from?” The pattern of coordinated deception becomes unmistakable. Consider how media outlets spent three years pushing Russiagate conspiracies, driving unprecedented social division while laying the groundwork for what would become the greatest psychological operation in history. Today, while the media floods us with Ukraine coverage, BlackRock positions itself to profit from both the destruction and reconstruction. The pattern becomes unmistakable once you see it—manufactured crises driving pre-planned “solutions” that always expand institutional control.

Mainstream media operates on twin deceptions: misdirection and manipulation. The same anchors who sold us WMDs in Iraq, promoted “Russia collusion,” and insisted Hunter Biden’s laptop was “Russian disinformation” still occupy prime time slots. Just as we see with RFK, Jr.’s HHS nomination, the pattern is consistent: coordinated attacks replace substantive debate, identical talking points appear across networks, and legitimate questions are dismissed through character assassination rather than evidence. Being consistently wrong isn’t a bug—it’s a feature. Their role isn’t to inform but to manufacture consent.

The template is consistent: Saturate media with emotional spectacles while advancing institutional agendas with minimal scrutiny. Like learning to spot a fake smile or hearing a false note in music, you develop an instinct for the timing:

Money and Power:

Medical Control:

Digital Control:

As these deceptions become more obvious, different forms of resistance emerge. The truth-seeking takes different forms. Some become deep experts in specific deceptions—documenting early treatment successes with repurposed drugs, uncovering hospital protocol failures, or exploring the impact of vaccine injuries. Others develop a broader lens for seeing how narratives themselves are engineered.

Keep reading

US Military Shoots Down 3 Mexican Cartel Drones With Laser

A laser system was used to shoot down three drug cartel drones in southern Texas, close to the Mexican border, the U.S. military said on Aug. 27.

The North American Aerospace Defense Command and U.S. Northern Command, in a joint statement, said an Army Multipurpose High Energy Laser (AMP-HEL) was used overnight on Aug. 25-26 in the Rio Grande Valley.

The Joint Task Force-Southern Border (JTF-SB) deemed the three drones to be hostile and posing a physical threat to U.S. military personnel and their partners in U.S. Customs and Border Protection (CBP).

Maj. Gen. Curtis Taylor, commander of JTF-SB, said the successful operation against the drones was a testament to the readiness, vigilance, and technological edge of his force.

US Will ‘Not Tolerate Hostile Surveillance’

“Cartel networks are increasingly employing unmanned aircraft systems to facilitate illicit human-smuggling and actively spy on our personnel and law enforcement partners,” Taylor said. “By integrating layered countermeasures alongside advanced directed-energy capabilities like this laser system, we have made it clear that we will not tolerate hostile surveillance.”

Ammon Blair, a former U.S.Border Patrol agent and senior fellow at the Texas Public Policy Foundation’s Secure & Sovereign Texas Initiative, told The Epoch Times in March 2025 that he recalled hearing cartel-operated drones overhead while working along the Texas border.

“You’d be running, and you’d have a drone right above your head and tracking every move you made,” Blair said at the time.

In February, the airspace around El Paso International Airport was closed for 7 hours for “special security reasons.”

The White House told The Epoch Times in an emailed statement at the time that Mexican cartel-operated drones had breached U.S. airspace and that the Department of War took action to disable them.

In other parts of Mexico, cartels have used drones to drop explosives in deadly attacks, but there is no suggestion the drones in the Texas incident were armed.

Keep reading

Flock CEO Garrett Langley Targeted as Address Shared Online

Critics of Flock Safety are circulating an address they say belongs to company founder and CEO Garrett Langley, marking an escalation in the widespread backlash the surveillance technology company is facing amid concerns over privacy and potential police misuse.

Numerous posts sharing an address purportedly belonging to Langley have circulated on X, with many users responding to a viral post that quoted Langley as saying Americans must “compromise” on their privacy.

Some shared what they said was a blurred image of Langley’s home on Google Street View, with one X user saying the “compromise apparently doesn’t apply to him.”

Newsweek has not verified the address and is not publishing or linking to it. Newsweek has contacted Langley for comment via an email to Flock Safety.

The renewed attention came after Langley recently appeared on Fox News’ Saturday in America, where he called for a “compromise” between public safety and privacy.

Keep reading

Harvard Professor Warns That AI Execs Are Pushing World Toward Factory Farming of Humans

We’re racing towards a future where AI enables the “factory farming of humans,” a Harvard professor warns — if tech elites get their way, at least.

In a new episode of The New York Time’s “Hard Fork” podcast, the historian Jill Lepore discusses her idea of an “artificial state” that’s supplanting the traditional nation state, and particular constitutional democracy, with one that erases human autonomy.

In this emerging world order, government is conducted “not by the consent of people but by machines that are making decisions,” she explained on the podcast. And those machines, of course, are owned by corporations.

We aren’t living in the artificial state quite yet, Lepore stresses, but it’s something that’s being built and realized around us. And crucially, it’s also a fantasy held by certain influential elites who believe their power should supersede the actual government and the voters who decide how it’s run.

“The rhetoric that comes out of this particular historical moment is really about the evanescence of modern liberal democracy, constitutional democracy,” she said. The elites pushing this rhetoric aren’t “interested in what the people want,” she added. “Because actually, what the people want is not to have AI and not to have data centers. So that’s the crisis.”

Lepore explores the idea in depth in her upcoming book “The Rise and Fall of the Artificial State.” She was inspired by a simple question: in our current world where you’re forced to talk to an AI to get answers, ranging from when you call in to ask about an appointment or talk to a company’s support chatbot on its website, “who decided this is a way we should be living?”

The artificial state is a sort of backsliding, in which more and more of the state’s functions are now being replaced by corporations that own and control our digital communications infrastructure. How we get our news, and information on elections and who to vote for, is all being mediated by these powerful groups — through technologies like the TV and radio, once, and now social media being the increasingly dominant medium.

The digital communication our lives are now structured around have a powerful isolating effect. In her 2024 piece for The New Yorker where she initially explored the idea that would later become her book, Lepore described the artificial state as the “factory farming of public life, the sorting and segmenting, the isolation and alienation, the destruction of human community.” 

The artificial state isn’t purely driven by AI, but it’s certainly being accelerated by it. People are asking ChatGPT about who to vote for, rather than talking to their neighbors, local officials, or experts. The political messages they’re seeing, meanwhile, are probably in part being written with AI. Decisions at every level are being made with the assistance of thinking machines. AI is also entrenching surveillance, processing the mountains of data that is being gathered on all of us.

One nefarious aspect of the pro-AI rhetoric coming out of Silicon Valley is that it frames AI as an inevitability. But Lepore argued that this brand of fatalism is a “bogus premise,” itself nestled on top of several more dubious propositions. One we hear endlessly is that regulation stifles innovation, so the government should take the handbrake off AI development. The other is that technology always advances democracy, which assures us that all of the very real threats to democracy AI poses in the present will magically sort themselves out.

Neither of those are true, Lepore says, and you have to look no further than how similar promises were made for technologies in the recent past. 

“So now, the AI people come back with the same argument that was made about the internet — the personal computer, the internet, and social media,” she said in the interview.

Keep reading

The UN and Red Cross Are Calling for Restrictions on AI Drones and Killer Robots

The United Nations and the Red Cross are warning that rapid advances in artificial intelligence (AI) and robotic weapons are bringing the world dangerously close to crossing a moral red line that has inspired dystopian science fiction for decades: the autonomous targeting of humans by deadly machines.

In an appeal to world leaders on August 25, the International Committee of the Red Cross and UN Secretary-General António Guterres renewed their call for a global treaty establishing restrictions on autonomous weapons systems. They warned that the technology behind AI targeting software and unpiloted drones has developed at record speed since the UN and Red Cross first called for such negotiations three years ago.

“At the same time, scientists and engineers developing these systems have themselves raised the alarm,” Guterres and the Red Cross said in a joint statement on August 25. “When the architects of this technology are calling for limits, States cannot afford to remain passive. This year must mark a turning point.”

In other words, the world must act now to rein in what are commonly known as “killer robots” or “slaughterbots” — unpiloted drones, robotic dogs, robot sentries armed with guns, AI targeting systems, and other digitally driven death machines that can identify, strike, and eliminate human beings on their own.

“You don’t even need drone operators; you can just type in a coordinate and not see the implications of what you are doing,” said Rory Fanning, an author, activist, and veteran of the U.S. war on Afghanistan, in an interview.

Keep reading

Sanders Vows to Introduce Legislation to Stop Flock’s “AI Mass Surveillance”

Sen. Bernie Sanders (I-Vermont) has pledged to introduce legislation aimed at reining in the mass adoption of AI-powered video surveillance via companies like Flock Safety, as the camera networks being erected in communities across the U.S. face fierce backlash.

Sanders said that the widespread adoption of AI-powered Flock cameras that monitor peoples’ every activity outside of their homes is edging the U.S. closer to a system of totalitarianism. The company says that its cameras are already in place in over 6,000 communities across the U.S.

“You may not know it, but you may already be being tracked by artificial intelligence,” Sanders wrote on Monday. “Flock, a surveillance company backed by Big Tech oligarchs like Marc Andreessen and Peter Thiel, has installed more than 120,000 AI-powered surveillance cameras across America.”

“What does that mean for you? It means that if Flock is in your community, every time you leave your house, a Flock camera records and tracks you,” Sanders went on. “If you are dropping your kids off at school, Flock knows. If you are going to a doctor, Flock knows. If you are going out on a date, Flock knows.”

Sanders said that he is soon introducing legislation to “stop Flock and AI mass surveillance.”

Keep reading