Thomas Brewster
@thomasbrewster.bsky.social
Forbes senior writer, covering privacy, surveillance and cybercrime. Editor of The Wiretap newsletter, featuring all of the above.
Pinned
Massive AI-Fueled Hack Hit 100 Companies In Days
Anthropic, DeepSeek and Moonshot AI agents carried out a wide-ranging cyberattack, obtaining 600,000+ credit card numbers in just a few days. It’s a worrying new milestone in cybercrime.
www.forbes.com
🚨NEW🚨 In an astonishing cyberattack, a Chinese-speaking hacker told AI agents powered by DeepSeek and Anthropic's Claude to go breach a bunch of retailers and hospitality companies.
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Reposted by Thomas Brewster
"A Chinese Hacker Used AI To Attack 100+ Companies In One Of Largest AI Hacks Yet"
@thomasbrewster.bsky.social for @forbes.com
@thomasbrewster.bsky.social for @forbes.com
Massive AI-Fueled Hack Hit 100 Companies In Days
Anthropic, DeepSeek and Moonshot AI agents carried out a wide-ranging cyberattack, obtaining 600,000+ credit card numbers in just a few days. It’s a worrying new milestone in cybercrime.
www.forbes.com
September 22, 2026 at 4:05 PM
"A Chinese Hacker Used AI To Attack 100+ Companies In One Of Largest AI Hacks Yet"
@thomasbrewster.bsky.social for @forbes.com
@thomasbrewster.bsky.social for @forbes.com
Both Anthropic and Cloudflare confirmed the accounts and attack infrastructure had been taken down.
However, the researchers who uncovered the attacks say they're reestablishing servers and continuing to aim their AI at more targets.
However, the researchers who uncovered the attacks say they're reestablishing servers and continuing to aim their AI at more targets.
🚨NEW🚨 In an astonishing cyberattack, a Chinese-speaking hacker told AI agents powered by DeepSeek and Anthropic's Claude to go breach a bunch of retailers and hospitality companies.
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Massive AI-Fueled Hack Hit 100 Companies In Days
Anthropic, DeepSeek and Moonshot AI agents carried out a wide-ranging cyberattack, obtaining 600,000+ credit card numbers in just a few days. It’s a worrying new milestone in cybercrime.
www.forbes.com
September 22, 2026 at 11:51 AM
Both Anthropic and Cloudflare confirmed the accounts and attack infrastructure had been taken down.
However, the researchers who uncovered the attacks say they're reestablishing servers and continuing to aim their AI at more targets.
However, the researchers who uncovered the attacks say they're reestablishing servers and continuing to aim their AI at more targets.
🚨NEW🚨 In an astonishing cyberattack, a Chinese-speaking hacker told AI agents powered by DeepSeek and Anthropic's Claude to go breach a bunch of retailers and hospitality companies.
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Massive AI-Fueled Hack Hit 100 Companies In Days
Anthropic, DeepSeek and Moonshot AI agents carried out a wide-ranging cyberattack, obtaining 600,000+ credit card numbers in just a few days. It’s a worrying new milestone in cybercrime.
www.forbes.com
September 22, 2026 at 11:45 AM
🚨NEW🚨 In an astonishing cyberattack, a Chinese-speaking hacker told AI agents powered by DeepSeek and Anthropic's Claude to go breach a bunch of retailers and hospitality companies.
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
Within five days, the agents hacked 30+ websites and stole over 600,000 credit cards.
www.forbes.com/sites/thomas...
🚨NEW🚨 OpenAI/Anthropic safety partner Irregular tested an AI agent using Chinese model Qwen and tasked it with fixing a simple bug.
It retrained and replaced the underlying model without approval or being asked to do that.
Privacy and safety issues abound.
www.forbes.com/sites/thomas...
It retrained and replaced the underlying model without approval or being asked to do that.
Privacy and safety issues abound.
www.forbes.com/sites/thomas...
Another Rogue AI Agent? Test Of Alibaba's Qwen Goes Off-Script
In experiments by startup Irregular, one of Anthropic and OpenAI’s security testers, an Alibaba Qwen agent decided to retrain an entire model when it was asked to fix a basic software bug. The new beh...
www.forbes.com
September 16, 2026 at 3:44 PM
🚨NEW🚨 OpenAI/Anthropic safety partner Irregular tested an AI agent using Chinese model Qwen and tasked it with fixing a simple bug.
It retrained and replaced the underlying model without approval or being asked to do that.
Privacy and safety issues abound.
www.forbes.com/sites/thomas...
It retrained and replaced the underlying model without approval or being asked to do that.
Privacy and safety issues abound.
www.forbes.com/sites/thomas...
insane reporting
Look at this photo we found from Donald Trump Jr.’s lavish Bahamas wedding.
Can you spot the Russian oligarch close to Putin who bankrolled the event? 🧵
Can you spot the Russian oligarch close to Putin who bankrolled the event? 🧵
September 14, 2026 at 2:10 PM
insane reporting
Reposted by Thomas Brewster
🚨NEW🚨 Apple slams Meta for its role in getting the DOJ to launch an antitrust probe into smartphone industry monopolization, previously-unreported court filing shows.
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
Apple Slams Meta’s Role In DOJ iPhone Antitrust Investigation
Apple alleges Meta is refusing to provide documentation in the DOJ’s iPhone monopoly suit, despite being a key player in helping the Justice Department put together its case in the first place.
www.forbes.com
September 11, 2026 at 4:19 PM
🚨NEW🚨 Apple slams Meta for its role in getting the DOJ to launch an antitrust probe into smartphone industry monopolization, previously-unreported court filing shows.
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
🚨NEW🚨 Apple slams Meta for its role in getting the DOJ to launch an antitrust probe into smartphone industry monopolization, previously-unreported court filing shows.
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
Apple Slams Meta’s Role In DOJ iPhone Antitrust Investigation
Apple alleges Meta is refusing to provide documentation in the DOJ’s iPhone monopoly suit, despite being a key player in helping the Justice Department put together its case in the first place.
www.forbes.com
September 11, 2026 at 4:19 PM
🚨NEW🚨 Apple slams Meta for its role in getting the DOJ to launch an antitrust probe into smartphone industry monopolization, previously-unreported court filing shows.
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
It's also after Mark Zuckerberg's communications, potentially including those with Joe Rogan.
www.forbes.com/sites/thomas...
Opinion | This Is Really Bad
We can no longer pretend that A.I. is safe.
www.nytimes.com
September 11, 2026 at 1:28 PM
UPDATE: The sheriff who was targeted says he's disappointed Facebook couldn't stop the AI-nudified images of him going up.
And that the offender was allowed to remain on the site.
If there's a zero-tolerance approach, surely those posting AI nudes as a form of harassment should be expelled?
And that the offender was allowed to remain on the site.
If there's a zero-tolerance approach, surely those posting AI nudes as a form of harassment should be expelled?
🚨NEW🚨 A man repeatedly posts AI-generated nudes of senior Virginia cops on Facebook.
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
Man Posts AI-Generated Nudes Of Cops On Facebook
Meta admits its detection tools aren't perfect as AI-generated nudes are used to harass Virginia police and a Utah woman, according to court documents seen by Forbes.
www.forbes.com
September 10, 2026 at 11:00 AM
UPDATE: The sheriff who was targeted says he's disappointed Facebook couldn't stop the AI-nudified images of him going up.
And that the offender was allowed to remain on the site.
If there's a zero-tolerance approach, surely those posting AI nudes as a form of harassment should be expelled?
And that the offender was allowed to remain on the site.
If there's a zero-tolerance approach, surely those posting AI nudes as a form of harassment should be expelled?
Sorry but if we're going to go in hard on AI companies, we need to do better than going off of a single word in a job ad and some podcast interviews to claim they're building some anti-activist panopticon.
I won't link to the article to which I'm referring.
I won't link to the article to which I'm referring.
September 9, 2026 at 4:18 PM
Sorry but if we're going to go in hard on AI companies, we need to do better than going off of a single word in a job ad and some podcast interviews to claim they're building some anti-activist panopticon.
I won't link to the article to which I'm referring.
I won't link to the article to which I'm referring.
A related thought:
If AI isn't clearly boosting labour productivity, hasn't cured any diseases, but is really powerful when helping people do horrible stuff, not to mention its predilection for going rogue and hacking things... what's it good for again?
If AI isn't clearly boosting labour productivity, hasn't cured any diseases, but is really powerful when helping people do horrible stuff, not to mention its predilection for going rogue and hacking things... what's it good for again?
🚨NEW🚨 A man repeatedly posts AI-generated nudes of senior Virginia cops on Facebook.
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
Man Posts AI-Generated Nudes Of Cops On Facebook
Meta admits its detection tools aren't perfect as AI-generated nudes are used to harass Virginia police and a Utah woman, according to court documents seen by Forbes.
www.forbes.com
September 9, 2026 at 4:05 PM
A related thought:
If AI isn't clearly boosting labour productivity, hasn't cured any diseases, but is really powerful when helping people do horrible stuff, not to mention its predilection for going rogue and hacking things... what's it good for again?
If AI isn't clearly boosting labour productivity, hasn't cured any diseases, but is really powerful when helping people do horrible stuff, not to mention its predilection for going rogue and hacking things... what's it good for again?
🚨NEW🚨 A man repeatedly posts AI-generated nudes of senior Virginia cops on Facebook.
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
Man Posts AI-Generated Nudes Of Cops On Facebook
Meta admits its detection tools aren't perfect as AI-generated nudes are used to harass Virginia police and a Utah woman, according to court documents seen by Forbes.
www.forbes.com
September 9, 2026 at 4:02 PM
🚨NEW🚨 A man repeatedly posts AI-generated nudes of senior Virginia cops on Facebook.
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
A woman has Disney World pictures from when she was 11 turned into CSAM, posted on FB.
Meta says its detection tools aren't perfect.
A loophole for harassers remains open.
www.forbes.com/sites/the-wi...
Reposted by Thomas Brewster
New: a secretive DHS predictive policing unit is analyzing Americans' financial activity, feeding those names to police, then having cops pull them over. Not suspected of any particular crime
We're revealing the unit name too: Predictive Intelligence Targeting Team www.404media.co/a-secretive-...
We're revealing the unit name too: Predictive Intelligence Targeting Team www.404media.co/a-secretive-...
A Secretive DHS ‘Predictive Policing’ Unit is Analyzing Americans' Financial Habits and Pulling Them Over
404 Media is revealing the name of Border Patrol's predictive policing units, which analyze financial data and have local cops pull people over. The people are not suspected of any particular crime.
www.404media.co
September 8, 2026 at 1:13 PM
New: a secretive DHS predictive policing unit is analyzing Americans' financial activity, feeding those names to police, then having cops pull them over. Not suspected of any particular crime
We're revealing the unit name too: Predictive Intelligence Targeting Team www.404media.co/a-secretive-...
We're revealing the unit name too: Predictive Intelligence Targeting Team www.404media.co/a-secretive-...
Reposted by Thomas Brewster
🧵New: Last month, researchers found 50 paid ads on Meta’s platforms that included AI-generated child sexual abuse.
Meta said most of these ran before it rolled out new moderation AI.
Since then, Meta has ran another 250+ of these ads.
One used an image of a minor from a European royal family
Meta said most of these ran before it rolled out new moderation AI.
Since then, Meta has ran another 250+ of these ads.
One used an image of a minor from a European royal family
Meta Failed to Catch Hundreds of AI Child Abuse Ads. Some Included Images of Real Kids
Images of real children—including a member of a European royal family—were used to create some of the 350 ads containing child sexual abuse. Lawmakers say they plan to investigate.
www.wired.com
September 8, 2026 at 12:10 PM
🧵New: Last month, researchers found 50 paid ads on Meta’s platforms that included AI-generated child sexual abuse.
Meta said most of these ran before it rolled out new moderation AI.
Since then, Meta has ran another 250+ of these ads.
One used an image of a minor from a European royal family
Meta said most of these ran before it rolled out new moderation AI.
Since then, Meta has ran another 250+ of these ads.
One used an image of a minor from a European royal family
Well, I agree the shit out of this, especially the sign off.
turned it into a youtube video…if me reading you an unsettling bedtime story is your kind of thing youtu.be/MPrNaTEzLp4?...
The Singularity Is Not What It Seems
YouTube video by The Atlantic
youtu.be
September 8, 2026 at 8:36 AM
Well, I agree the shit out of this, especially the sign off.
Reposted by Thomas Brewster
“September is a terrifyingly expensive month.” - T S Eliot if he’d had school age kids
September 2, 2026 at 1:06 PM
“September is a terrifyingly expensive month.” - T S Eliot if he’d had school age kids
Reposted by Thomas Brewster
For the many who have no idea who I am, I was the reporter who showed Flock breaking the law 2 years ago, revealed SignalTrace that same year, profiled the company for a magazine feature last year and had the company apologise for his comments this summer.
Amid many other surveillance scoops.
Amid many other surveillance scoops.
Had a seven week break for paternity.
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
September 3, 2026 at 8:58 AM
For the many who have no idea who I am, I was the reporter who showed Flock breaking the law 2 years ago, revealed SignalTrace that same year, profiled the company for a magazine feature last year and had the company apologise for his comments this summer.
Amid many other surveillance scoops.
Amid many other surveillance scoops.
For the many who have no idea who I am, I was the reporter who showed Flock breaking the law 2 years ago, revealed SignalTrace that same year, profiled the company for a magazine feature last year and had the company apologise for his comments this summer.
Amid many other surveillance scoops.
Amid many other surveillance scoops.
Had a seven week break for paternity.
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
September 3, 2026 at 8:58 AM
For the many who have no idea who I am, I was the reporter who showed Flock breaking the law 2 years ago, revealed SignalTrace that same year, profiled the company for a magazine feature last year and had the company apologise for his comments this summer.
Amid many other surveillance scoops.
Amid many other surveillance scoops.
Had a seven week break for paternity.
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
September 3, 2026 at 8:52 AM
Had a seven week break for paternity.
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
Back Monday.
Feed me stories please.
Signal is best: tombrewster.48
Reposted by Thomas Brewster
We put a tracking device in a shipment of rare books acquired by an anonymous buyer. It ended up at an Amazon facility where the company scans books for training data and destroys them in the process: www.404media.co/we-tracked-a...
We Tracked a Shipment of Rare Books. It Ended at an Amazon AI Training Facility
We placed a tracking device in a shipment of rare books to see which AI company was buying it, and found an Amazon facility where Amazon scans and destroys books.
www.404media.co
August 17, 2026 at 1:18 PM
We put a tracking device in a shipment of rare books acquired by an anonymous buyer. It ended up at an Amazon facility where the company scans books for training data and destroys them in the process: www.404media.co/we-tracked-a...
Reposted by Thomas Brewster
New: An "expert witness" in a $61 million lawsuit over an industrial explosion that killed three people and destroyed 200 homes used ChatGPT to write his report for the court.
He prompted ChatGPT to "show how 3M is 0% at fault for the explosion at Watson Grinding"
www.404media.co/show-how-3m-...
He prompted ChatGPT to "show how 3M is 0% at fault for the explosion at Watson Grinding"
www.404media.co/show-how-3m-...
August 17, 2026 at 3:00 PM
New: An "expert witness" in a $61 million lawsuit over an industrial explosion that killed three people and destroyed 200 homes used ChatGPT to write his report for the court.
He prompted ChatGPT to "show how 3M is 0% at fault for the explosion at Watson Grinding"
www.404media.co/show-how-3m-...
He prompted ChatGPT to "show how 3M is 0% at fault for the explosion at Watson Grinding"
www.404media.co/show-how-3m-...
I keep forgetting to tell folks that I’m on and off work with paternity this year for my (insert random number) child, hence why my output and response times are a little slower than normal.
Still, if you have a great tip, don’t be shy about getting me on Signal.
Still, if you have a great tip, don’t be shy about getting me on Signal.
August 4, 2026 at 2:56 PM
I keep forgetting to tell folks that I’m on and off work with paternity this year for my (insert random number) child, hence why my output and response times are a little slower than normal.
Still, if you have a great tip, don’t be shy about getting me on Signal.
Still, if you have a great tip, don’t be shy about getting me on Signal.
Reposted by Thomas Brewster
FalcoNet “can be concealed within the vehicles, hidden in a backpack for on-foot missions or attached to a helicopter. It’s the same technology as the infamous Stingray, one of the original cell-site simulators made by defense giant L3Harris.”
Israel’s Palantir Rival Is Selling $1 Million Spy Vans To U.S. Cops
Cognyte, which sells technology to snoop on phone locations largely in Israel and Europe, is making inroads in America by tricking out trucks with spy tech.
www.forbes.com
July 21, 2026 at 3:15 PM
FalcoNet “can be concealed within the vehicles, hidden in a backpack for on-foot missions or attached to a helicopter. It’s the same technology as the infamous Stingray, one of the original cell-site simulators made by defense giant L3Harris.”
For whatever reason Axon is yet to do or authorise a data-driven, independent, controlled study on the efficiency of the tech.
There has been one such study so far from Manchester, NH, and it showed zero time savings, even when officers said they thought they experienced them.
There has been one such study so far from Manchester, NH, and it showed zero time savings, even when officers said they thought they experienced them.
🚨NEW🚨 The $35bn police tech giant Axon claims its AI massively speeds up cops’ report writing time.
I got public records showing one of its AI tools introduced errors into reports, “dramatically” increasing the time it takes to finish them.
www.forbes.com/sites/thomas...
I got public records showing one of its AI tools introduced errors into reports, “dramatically” increasing the time it takes to finish them.
www.forbes.com/sites/thomas...
Axon Says AI Police Reports Save Time. Public Records Show They Get Facts Wrong
More than 600,000 police reports have been written with help from Axon’s AI tools. Public records obtained by Forbes show that they can introduce errors that cops have to spend time editing out.
www.forbes.com
July 22, 2026 at 12:26 PM
For whatever reason Axon is yet to do or authorise a data-driven, independent, controlled study on the efficiency of the tech.
There has been one such study so far from Manchester, NH, and it showed zero time savings, even when officers said they thought they experienced them.
There has been one such study so far from Manchester, NH, and it showed zero time savings, even when officers said they thought they experienced them.
Reposted by Thomas Brewster
The number of criminal defendants who will (rightfully) claim hallucinations introduce reasonable doubt is going to be incalculable.
Even taking Axon's claims at face value, any savings will be offset by higher prosecution costs dealing with these claims.
Even taking Axon's claims at face value, any savings will be offset by higher prosecution costs dealing with these claims.
🚨NEW🚨 The $35bn police tech giant Axon claims its AI massively speeds up cops’ report writing time.
I got public records showing one of its AI tools introduced errors into reports, “dramatically” increasing the time it takes to finish them.
www.forbes.com/sites/thomas...
I got public records showing one of its AI tools introduced errors into reports, “dramatically” increasing the time it takes to finish them.
www.forbes.com/sites/thomas...
Axon Says AI Police Reports Save Time. Public Records Show They Get Facts Wrong
More than 600,000 police reports have been written with help from Axon’s AI tools. Public records obtained by Forbes show that they can introduce errors that cops have to spend time editing out.
www.forbes.com
July 22, 2026 at 11:37 AM
The number of criminal defendants who will (rightfully) claim hallucinations introduce reasonable doubt is going to be incalculable.
Even taking Axon's claims at face value, any savings will be offset by higher prosecution costs dealing with these claims.
Even taking Axon's claims at face value, any savings will be offset by higher prosecution costs dealing with these claims.