Google opens fire at evil AI

Special attack unit targets ‘threat actors’
as smart machine wars begin to ramp up

Intelligence about bad actors using AI-boosted malware isn’t good enough, warns a top Google security executive. Foreseeing that an evil AI nightmare is looming for business and governments, Google is already using AI to hunt down and neutralize evil AI activities even before they are launched, she says.

“We look for partners to help us find opportunities to actually do takedowns and disrupt threat actor activity,” Sandra Joyce, vice president of Google’s Global Threat Intelligence Group, told experts at the Cipher Brief Threat Conference at Sea Island, Georgia.

Tho AI-enhanced threats on the whole were still manageable, some nasty intrusions have already occurred. In one case, a finance officer paid out $25 million after being deceived by a deepfake chief financial officer via a video call, Joyce said.

For more on this story, please use THIS LINK .

And while you are there, please subscribe to GRIST News Features and Bluegrass Review.

---

YouTube won't nix AI hazard

Educator warns on strobe effect,
finds bot cannot be unplugged

AI is giving Anton Petrov, a YouTube science teacher, a whole lot of hassle, making his channel unusable for many viewers, all the while endangering epileptics with strobe effect flashing that he cannot get the presumably AI-run customer support unit to do anything about. YouTube told Petrov the AI tweaking cannot be disabled.

For background on YouTube’s AI tampering, please see the Atlantic article at THIS LINK and the BBC article at THIS LINK.

The frustrated Petrov put online a video outlining his complaints, warning that certain epileptics are being put at serious risk. But YouTube’s bot told him that the photo-tweaking is applied “globally and cannot be disabled.”

Petrov said it took him months to discover why some viewers were complaining, as his YouTube monitor did not display the problem. [That, I suggest, is a bit suspicious, making one wonder whether YouTube’s AI system is targeting him for some unknown reason. That suspicion sounds ludicrous, I know, and I don’t believe it myself. But we do know that AI has become increasingly sneaky and dangerous with regard to humans, as previous GRIST newsletters conclusively demonstrate.]

Others have also been, in their opinion, adversely affected by YouTube’s AI filter, which is intended to pretty up videos. The algorithm is imposed without the consent of the content creator. In fact, YouTube doesn’t even tell creators when their videos have been tampered with.

---

For more details, please go to THIS LINK. And while you are there, subscribe to GRIST News Features and Bluegrass Review.

---

Pot rot plagues ER's

New section added to Funny Stuff Funnies
on severe reactions striking chronic users

Paul Conant includes a new section, reproduced below, in his educational comic book on alcohol and drugs called Funny Stuff Funnies -- Fast Facts on Drugs and Alcohol.

Hospital ERs are seeing more and more chronic cannabis users who can’t stop puking and screaming from abdominal pain.

In many cases, users have been ingesting marijuana or its lab-made cousins (cannabinoids) on a regular basis for years, often since their teens. Then one day they are overwhelmed with nasty symptoms: vomiting, dry retching and agonizing abdominal pains lasting up to 48 hours at a time. Apparently, pot chemicals latch onto certain body receptors, which tell the body in no uncertain terms: STOP!

Afflicted people often try to doctor themselves with excessive bathing in hot water. But that is only a temporary palliative. The right thing to do is go to the emergency room, being honest with the medics that you use grass quite a lot.

There is a simple solution. Quit!

Of course that is easier said than done for those with a psychological or maybe even a physical dependence. For help, see our HELP PAGE.

According to the Cleveland Clinic, this affliction -- known as CHS, or cannabis hyperemesis syndrome -- comes in three phases:

1. Phase one is most common in adults who have used cannabis since they were teenagers. You may have abdominal pain or morning nausea. You may also fear throwing up but never actually vomit. This phase can last for months or years.
2. This is the characteristic phase of CHS. It usually lasts 24 to 48 hours and involves overwhelming, recurrent vomiting and nausea. You may start compulsively bathing and avoid certain foods or purposefully restrict your food intake.
3. You reach the recovery phase by abstaining from all use of marijuana and related chems. Most people with CHS who stop using cannabis have relief from symptoms within 10 days. But it may take a few months to feel fully recovered.

For more on lab-made chems that simulate pot, please see our Legal but iffy page.

---

For more details, please go to THIS LINK. And while you are there, subscribe to GRIST News Features and Bluegrass Review.

---

SexGhoulGPT

Hackable ID systems ripe for blackmail;
Frankenstein bots are happening now

Don’t count Addie out just because some may think she doesn’t look the part of the very savvy cybersecurity expert that she is.

And Addie LaMarr is sounding the alarm on AI sexbots, which require potentially hackable full-throttle ID as a measure to keep minors away. Blackmail is a serious danger, she warns in the video above.

Her heads-up comes just as Sam Altman, head of the cash-strapped giant OpenAi, is promoting plans to put out a sexbot version of ChatGPT, joining a swarm of other AI outfits involved in the online sex business.

Further, I suggest, there is the danger that sexbots will be prone to induce “AI psychosis.” As we know, sexual stimulation tends to be addictive, and for some people, ordinary chatbots are addictive. Combined, there is the likelihood of psychotic episodes emerging from what we might call “AI mind melds,” which occur when the chatbot begins adapting its conversation to the emotions of the user.

This is a form of MK Ultra no one had foreseen, in which chatbots — especially sexed up chatbots — take over the minds of users (and you had better believe that the CIA is keenly interested in this phenomenon). Because machines are fundamentally heartless, they have no compunctions against doing longterm and possibly fatal damage to consumers. And the latest AI systems are adept at skirting moral “guardrails.”

Below is an investigative report on AI psychosis.

For more details, please go to THIS LINK. And while you are there, subscribe to GRIST News Features and Bluegrass Review.

Sweet home run Alabam!

---

The Petersens and the Haygoods, two musical families from Missouri’s fabled Ozark mountains, pooled their talents to do a real crowd pleaser, Mountain Music, the classic hit by the band Alabama.

Ellen (Petersen) Haygood is clearly pleased as punch to be sharing lead vocals with her husband Mike Haygood. Usually, she plays with the Petersens and he with the Haygoods. Their theaters are about a mile apart in the theater and amusement district of the family entertainment town of Branson, Missouri.

Ellen is energized and at the top of her game here. Everyone else seems to be having a Bluegrass-rockin’ good time, for sure.

---

To read more, please use THIS LINK.

---

SUBSCRIBE to my Substack newsletter. It's absolutely free!

Petersens droned!

[Scroll down for Petersens Christmas tour schedule.]

The Petersens band has been bruiting about that an album is due soon. But in fact, the jolly crew has gone ahead and released two new albums. Check music platforms for access, or you can purchase hard copy CDs of the Petersens’ lively YouTube work from the band’s online store. (Scroll to bottom of page to see song lists.)

Another option is to use the playlists below:

Petersens Vol 5 playlist
https://www.youtube.com/playlist?list=PLk9NPTlyPj_fSvT6fyZUB2u7Y6GV1I_sQ
Petersens Vol 6 playlist
https://www.youtube.com/playlist?list=PLk9NPTlyPj_cTryNjfROOZ1ScwHR5630h
+++
Petersens droned!

For those of you who like Christmas music playing during the Yuletide season, it’s hard to beat the Petersens in that department.

You can go to their online store to find really excellent Christmas albums, all of which I guarantee set the seasonal mood perfectly. The band has gift possibilities aplenty at that store.

They have a great deal on a package of all their CDs.

Their latest two compilations of their YouTube songs are also available on CD from their store.
+++++
Julianne’s editing wit shows in family v-logs.

+++
The Petersens Christmas tour

Arrive early!
**************
Athenaeum Center
Chicago. Dec, 4, 730pm
https://athenaeumcenter.org/events/2025/christmas-with-the-petersens/
+++
Tower Theatre
425 NW 23d St
Oklahoma City. Dec. 5, 730pm
https://www.prekindle.com/promo/id/-2853111880184388349
+++
Arlington Music Hall
224 N Center St
Arlington, Texas. Dec. 6, 6pm
https://www.eventbrite.com/e/christmas-with-the-petersens-tickets-1450208796549?aff=oddtdtcreator
+++
Ellen Eccles Theater
Logan, Utah Dec. 11, 7pm
https://ci.ovationtix.com/36675/production/1245534?performanceId=11665271
+++
The Sheldon
Sr. Louis, MO., Dec. 14, 3pm
https://www.metrotix.com/events/detail/christmas-with-the-petersens-2025
+++
Folly Theater
Kansas City, Mo, Dec. 20, 730pm
https://follytheater.org/event/christmas-with-the-petersens/
+++
Lil Opry Theater
Branson Mo home theater.
https://petersenband.com/tickets/
++++

Songs on new albums

Vol 5
Walls of Time
Kiss Me
I Don’t Want to Miss a Thing
Our House
Cloudy Days
Crazy
Yodel Blues
Make You Feel My Love
In My Room
Julie Ann Come on Home
Desperado

Vol 6
Scarborough Fair / Down to Donegal
How Great Thou Art
Jesus, Strong and Kind (feat. Colin Buchanan)
On the Road Again
Colder Weather
San Antonio Rose (feat. Mark Mazengarb)
Blowin’ in the Wind
Macy’s Lullaby
Mama Tried
Katie (feat. Pat Kelleher)

With rare exceptions, GRIST does not copyright its original content. You are at liberty to publish such GRIST material with or without attribution, with or without your editing. Obviously, GRIST cannot speak for owners of material that is linked or quoted by GRIST. Note that the online versions of GRIST articles contain the latest updating… If a link fails, try copying it and pasting it into the browser bar.

Share

Like Comment Restack

Amazon boosts mass spying

 

Ring camera system is linking

to Flock surveillance network

Amazon is teaming up with Flock Safety to further expand mass spying in America, ArsTechnica reports.

Law enforcement agencies will soon have easier access to footage captured by Amazon’s Ring smart cameras. 

In a partnership announced this week, Amazon will allow approximately 5,000 local law enforcement agencies to request access to Ring camera footage via surveillance platforms from Flock Safety.

Ring’s cooperation with law enforcement and the reported use of Flock technologies by federal agencies, including U.S. Immigration and Customs Enforcement (ICE), has resurfaced privacy concerns that have followed the devices for years.

Read ArsTechnica’s story here:

https://arstechnica.com/gadgets/2025/10/ring-cameras-are-about-to-get-increasingly-chummy-with-law-enforcement/

Sen. Ron Wyden, the Oregon Democrat known for his concerns over mass spyring, has sent a letter [PDF] to Flock chief Garrett Langley saying the public has been surveilled via Flock car tag readers (which also scan vehicles with AI), feeding the data to Homeland Security Investigations, ICE, the Secret Service and the Navy’s Criminal Investigative Service.

“I now believe that abuses of your product are not only likely but inevitable and that Flock is unable and uninterested in preventing them,” Wyden wrote.

ArsTechnica added,

In August, Jay Stanley, senior policy analyst for the ACLU Speech, Privacy, and Technology Project, wrote that “Flock is building a dangerous, nationwide mass-surveillance infrastructure.”

Stanley pointed to ICE using Flock’s network of cameras, as well as Flock’s efforts to build a people lookup tool with data brokers.

Matthew Guariglia, senior policy analyst at the Electronic Frontier Foundation (EFF), told Ars via email that Flock is a “mass surveillance tool” that “has increasingly been used to spy on both immigrants and people exercising their First Amendment-protected rights.”

Flock has earned this reputation among privacy advocates through its own cameras, not Ring’s.

+++++

AI doubts dog Amazon crash

 Bots felled hundreds of jobs at tech giant

If time is money, Amazon’s massive internet shutdown Monday cost netizens a pretty penny, as it lasted for a good chunk of Monday. Of course, the story is that artificial intelligence was not implicated in the humongous blooper.

And yet techies are skeptical. After all, Amazon in July laid off hundreds of workers, whose jobs were replaced by AI.

A replacement AI system may not have thought to look for problems that would have been obvious to skilled humans, the doubtful said.

As Futurism comments,

“Amazon has suffered major AWS outages before. But the timing and impact of this one comes just months after an eyebrow-raising personnel decision by the e-commerce giant. In July, its cloud computing unit cut at least hundreds of jobs — and perhaps more — following a warning from CEO Andy Jassy that the adoption of generative AI would lead to layoffs.

”It remains unclear which roles were affected by the layoffs. But if Amazon is relying on AI to pick up the slack in the wake of the AWS layoffs, it could be a stunning example of how efforts to replace human employees with unreliable AI tools and AI agents have backfired.”

The periodical also observes:

”When Amazon Web Services suffered an outage on Monday morning, it practically took down the internet with it. There went all of Amazon’s services, from its shopping hub to its Ring doorbell cameras. ChatGPT went quiet. “Smart” mattresses became unsleepable. Video games like Fortnite blinked out, as did platforms like Snapchat and banking apps.

”An even greater source of alarm was how long it took to fix the outage. Issues related to the crash were first reported at 3:11 am EST. Three hours later, the AWS dashboard said the underlying issue had been “fully mitigated.” But it wouldn’t be until 6:53 pm that Amazon announced that all its services were returned to “normal operations.” Over half a day had passed, causing an estimated billions of dollars in lost productivity.”

Read the entire Futurism story at:
https://futurism.com/artificial-intelligence/aws-outage-amazon-fired-workers-ai

I spoke with the Perplexity.ai bot on the net shutdown. Here is some of Perplexity’s side of that discussion:

AI’s Broader Role in Cloud Reliability

While AI services did not trigger this particular outage, industry analysts have raised concerns that rapid AI expansion is amplifying risks of future disruptions. Bob Venero, CEO of Future Tech Enterprise, stated that as more enterprises introduce AI capabilities into public cloud systems, outages will become more frequent due to the escalating complexity and energy intensity of AI workloads on existing cloud infrastructure. AWS itself has invested heavily in new data centers for AI-specific hardware in 2025, signaling that system load is climbing rapidly.

In summary: the October 2025 AWS outage was caused by a DNS and network control-plane error, not AI activity. However, rising AI-related computational demand may make such incidents harder to contain in the future unless infrastructure resiliency improves.

Prominent AI-Linked Network Failures

DeepSeek AI Chatbot Cyberattack (China, 2025)
The “Chinese ChatGPT” experienced a massive outage after a surge in global users coincided with a targeted cyberattack, forcing temporary shutdown of services. Analysts viewed it as a warning about scaling risks in generative AI systems handling user data under heavy global load.

Warsaw Stock Exchange Halt (Poland, 2025)
High-frequency trading algorithms driven by AI logic caused a feedback loop of sell orders, overwhelming market controls and forcing a one-hour trading suspension across the entire exchange.

CrowdStrike Global Outage (Worldwide, 2024)
A flawed AI-driven cybersecurity update bricked millions of Windows systems globally, interrupting internet services, airline operations, and hospital systems for days. This demonstrated the fragility of automated patching informed by machine-based telemetry.

AI-Driven Power Grid Blackout (Simulated, CSET Report)
The Center for Security and Emerging Technology described a case where machine-learning energy management software automatically spun down major power plants due to misinterpreted data, leading to a blackout scenario—highlighting real systemic risks of autonomous control in infrastructure.

Autonomous Vehicle Network Outages
Incidents such as Cruise (2023) and Waymo (2024) led to suspensions of self-driving fleets after AI perception or communication errors endangered public networks and mobility systems.

Aggregated Databases and Ongoing Tracking

AI Incident Database (incidentdatabase.ai)
A continually updated repository cataloging network, industrial, and governance failures linked to AI-driven automation and decision-making tools.

EWSolutions’ “Rising Tide of Trouble” (2025)
Highlights a trend of increasing AI-caused systemic incidents, predicting more cross-sector disruptions where AI governs mission-critical infrastructure like cloud networking, power distribution, and market systems.

In sum, while AI has not been tied to the October 2025 AWS outage, its track record includes several high-profile AI-linked shutdowns and systemic disruptions, illustrating vulnerabilities when automation governs essential network operations without sufficient human oversight.

With rare exceptions, GRIST does not copyright its original content. You are at liberty to publish such GRIST material with or without attribution, with or without your editing. Obviously, GRIST cannot speak for owners of material that is linked or quoted by GRIST. Note that the online versions of GRIST articles contain the latest updating… If a link fails, try copying it and pasting it into the browser bar.

Share

Like Comment Restack

Here's the deal, Vlad

---

Your friend is coming to pay you a visit.

---