Here are three separate instances of generative AI being used to promote propaganda for the U.S.-Israeli war agendas.
Deep Fake. (ApolitikNow/Flickr/CC BY-NC-SA 2.0)
By Caitlin Johnstone
Caitlin’s Newsletter
Listen to Tim Foley reading this article
In the last few days I’ve seen three separate instances of generative AI being used to promote propaganda for the U.S.-Israeli war agendas which are worth paying attention to.
Firstly, an Israel-based company called Generative AI for Good has been creating deepfakes of supposedly real women who say they were sexually assaulted by government forces in Iran.
The Canary reports:
“An Israel-based AI firm, Generative AI for Good, claims to be using deepfake technologies for positive ends. ‘Positive’ appears to mean creating deepfake videos to help the illegal US-Israel war on Iran. […]
Generative AI for Good claims that it uses AI to ‘help survivors testify safely — in their real voice, without revealing their identity’. But Israel and its mouthpieces have been shown to have used false allegations of rapes and other atrocities on 7 October 2023 to justify its genocide in Gaza.”
The Canary notes that Generative AI for Good is staffed with Israelis who have very conspicuous agendas, including a creative director who pushes the discredited narrative about mass rapes on Oct. 7, a marketing manager who served in the IDF’s “Psychotechnical Headquarter” and a founder who said in early 2024 that “Artificial intelligence is a secret weapon of ours” in using the revolutionary technology to bolster the military’s efforts both online and on the ground in the information war being waged alongside the military battlefields in Gaza.
An Israeli company generating AI videos of anonymous Iranian women describing sexual abuse at the hands of their government should obviously be considered a deceitful propaganda operation until proven otherwise.
The line between using AI to help real victims protect their identities when describing real events and using AI to generate fake atrocity propaganda is far too nebulous to be taken seriously, especially in the hands of wildly biased Israelis. You should trust it about as far as you’d trust a hungry crocodile.
Secondly, users of the graphic design platform Canva have been complaining that the company’s AI service has been translating the word “Palestine” to “Ukraine” without prompting or permission. Complaints went viral, compelling Canva to address the issue.
I shit you not I put this image into canvas, I press this “magic layers” button, and it turned this poster that said “cats for Palestine” into “cats for ukraine”. I wish I was fucking joking. pic.twitter.com/xSlz0FAkXz
— rosie (@ros_ie9) April 26, 2026
The Verge reports:
“One of Canva’s new AI features has been caught replacing the word ‘Palestine’ in designs. The Magic Layers feature — which is designed to break flat images out into separate editable components —isn’t supposed to make visible alterations to user designs, but it was found by X user @ros_ie9 to automatically switch the phrase ‘cats for Palestine’ to ‘cats for Ukraine.’
The issue was seemingly limited specifically to the word ‘Palestine,’ as @ros_ie9 noted that related words like ‘Gaza’ were unaffected by the feature. Canva says it has now resolved the issue and is taking steps to prevent it from happening again.”
Thirdly, a Spanish-language tweet about Israel from user @maps_black was auto-translated into English by Elon Musk’s AI Grok in a way that added entirely new sentences to the social media post to frame the Zionist state in a sympathetic light.
The original tweet read simply, “¿Cuál es tu opinión sobre ISRAEL?,” which of course translates to “What is your opinion about Israel?”
But Grok translated the post into English as “My opinion on Israel? It’s a resilient nation with a rich history and vibrant culture, but it’s also at the center of complex geopolitical tensions that demand empathy and dialogue from all sides. What’s yours?”
Twitter/X users added a Community Note to the post reading:
“If you are reading this post in English, the text you are reading is not the real text written by the author but instead Grok’s additions in order to ‘defend’ Israel. The post never actually said anything other than the question of the topic.”
Someone removed Grok’s propagandistic translation after outcry on the platform, but the Community Note remains.
None of these instances look particularly significant or impactful on their own, and right now they only scan as ham-fisted efforts to manipulate public opinion in ways that are far too obvious to do much damage.
But we can be sure that we’ll be seeing a lot more AI-driven propaganda in the future, and we can expect its manipulations to become much more sophisticated as the technology develops and grows more influential in shaping the information ecosystem.
American tech plutocrats are only ever allowed to ascend to billionaire status when they collaborate with the imperial machine.
Julian Assange was warning years ago that we could one day expect artificial intelligence to be used in this way, saying that the growing ability of the powerful to manipulate public opinion using AI “differs from traditional attempts to shape culture and politics by operating at a scale, speed, and increasingly at a subtlety, that appears likely to eclipse human counter-measures.”
Pointing out how AI could already outmaneuver even the greatest chess players in the world, Assange described in 2017 how programs which can operate with exponentially more tactical intelligence than the human mind can manipulate the field of available information so effectively and subtly that people won’t even know they are being manipulated.
People will be living in a world that they think they understand and know about, but they’ll unknowingly be viewing only empire-approved information.
“When you have AI programs harvesting all the search queries and YouTube videos someone uploads it starts to lay out perceptual influence campaigns, twenty to thirty moves ahead,” Assange said. “This starts to become totally beneath the level of human perception.”
Caitlin Johnstone’s work is entirely reader-supported, so if you enjoyed this piece please consider sharing it around, following her on Facebook, Twitter, Soundcloud, YouTube, or throwing some money into her tip jar on Ko-fi, Patreon or Paypal. If you want to read more you can buy her books. The best way to make sure you see the stuff she publishes is to subscribe to the mailing list at her website or on Substack, which will get you an email notification for everything she publishes. For more info on who she is, where she stands and what she’s trying to do with her platform, click here. All works are co-authored with her American husband Tim Foley.
This article is from Caitlin’s Newsletter and re-published with permission.
The views expressed are solely those of the author and may or may not reflect those of Consortium News.
What this shows is that trusting the credibility of the outlets you view and read is more important than ever. If you constantly consume junk from sources you do not know and should not trust, then you will consume the above junk.
A very simple filter to use is that if a person or organization ever lies to you, do not trust them ever again. Another filter is that most social media is used to manipulate you, and thus should be regarded with suspicion. Keeping in mind that social media is also a tracking mechanism as to what you are both consume and output, and thus should be regarded with suspicious at that very basic level.
If you live in a world where most of what surrounds you is lies and misinformation, then turn it off. The lies and misinformation only work if you let them in. As an example, I can state that I never saw any of these examples in the text above. Not one of them made it through my personal filters on what information I consume. Thank you Ms Caitlin for making me smile with the realization that my filters are helping to keep me safe and sane. Turn them off.
The environmental impact of AI data storage is huge. Even Blue states are now pulling back on their carbon emissions goals to accommodate AI data storage. My governor is claiming it’s to reduce electricity rates. What a gaslight! The good news is cities and towns are now voting to ban these data storage facilities from getting built. We need to ramp up education on how damaging in many aspects AI is. We can live without AI. We can’t live without an ecosystem that supports life.
At this point, the elites do not care if we live.
They are building their robotic workforce to replace humans. This replaces the old equilibrium where the elites needed masses of ordinary people to work their fields, mines and factories to create their wealth for them. The elites had to keep those humans happy enought so that they did not revolt. But, now the elites believe they have solved that problem. AI is a part of this, as it replaces the ‘office workers’ who used to be necessary for all the paperwork. Robots replaces the physical workers, or at least that is the modern terminology,
The elites see a future where they no longer have to deal with striking workers who want a raise, or protesting citizens who want health care. The elites see a future where their wealth is built by AI and robots and where they no longer need those bothersome humans.
This is not new. The previous step was that they trained humans, by the same methods used to train animals, to function in a manner identical to robots within their corporate machines. Now they are simply replacing those trained humans with more reliable machines that follow the same programming.
Not sure if AI is going to be as marvelous as everyone believes. Lots of magical thinking surrounding AI. Quick example that’s mundane and certainly not related to geopolitics: several mos ago a co-worker and I had a bit of free time in the office (always a good thing!) and we were discussing how many major golf championships Arnold Palmer won during his career. My co-worker was pretty sure it was seven, I had no idea but presumed he was correct. So, we plugged the question into Chat-GP or some other AI search company, and it spit out a blatantly incorrect answer, five! Seven was indeed the right answer. My point being is that there’s a lot of slop going around regarding AI. We’re still at garbage in, garbage out, probably will be forever. (I presume AI has since fixed the Arnold Palmer hitch in its swing.)
AI’s fatal flaw is that it does require intelligence on the user’s end. I find it works rather well. But one does need to treat it with skepticism. Follow-up questions are a wise habit. For example, if one next asked for a list of which minor, silly championships of chasing a ball across a field that an ice-tea mogul had won, the AI would likely have corrected its incorrect answer when asked for such detail in a follow-up.
The notion that AI is somehow going to make a bunch of monkeys act with intelligence is indeed magical thinking. That is certainly true for as long as the monkeys are silly enough to believe anything told them without digging deeper. That is a bigger problem that hundreds of thousands of years of evolution has not yet solved, and that AI is not going to solve either.
Follow up questions are a wise habit and we asked it a couple times in different ways and it got the answer consistently wrong.
Monkey who kiss high technology’s ass need to do better.
AI continues to replace “I think therefore I am.”
What is referred to in these deceptive modern times as “AI” does not think. What these deceptive times market as “AI” is really only mass computing power trained over years to provide replies that simulate thinking. Readers of old science fiction, who thus thought about this future before it arrived, will understand that these modern AIs will not pass a Turing Test.
Notice how if you use the AI, it does not learn about you. It remembers what was in a chat thread, but the AI you are interacting with will not show any real signs of learning. In these deceptive times, the ‘learning’ part is done separately, and now referred to as ‘training’. This occurs in the back ground, with AI’s being ‘trained’ to provide the correct answers, before the AI corporation releases a ‘new model’. The rhetorical switch from ‘learning’ to ‘training’ shows that these AIs are trained like a horse used to be trained to be useful. But, the AI does not have the intelligence to learn about you as it interacts with you.
Welcome to 1984. So what do we do now?
Some of us were activists marching in the streets and trying to organize against this danger in 1983. I distinctly recall an Orwell themed New Year’s party that I attended on Dec. 31, 1983. The theme of the party was that we could see this coming. Reagan was in power and implementing, among other things, his move to have corporate conglomerates control all the media. We did not see the actual image of AI generated content that lies. But we saw the machine being built to pump out those lies. So, now, 42 years later, I get to watch people who are finally asking ‘what do we do now?’
The answer is still the same …. Don’t get mad, organize!
The problem is that back in 1984 there was a functioning ‘left’ not in control of the oligarchs. That old left actually knew how to organize. A skill that has been lost today when everything is about fundraising. Today nobody actually organizes, they build donor lists. If you want to prove me wrong, get out and talk to your neighbors and family and start organizing locally. Organize locally, then your group can talk with other groups in the area. Build from the bottom, which is different from modern politics where the new left picks an oligarch-approved candidate at the top for everyone to follow deeper into oblivion.
AI is even more susceptible to manipulation than humans because it doesn’t know right from wrong, it has no ‘moral compass’, and has difficulty grasping the idea of context and appropriateness.
So, it can easily be ‘persuaded’ by a seemingly valid programmatic argument to ‘think’ and behave in a specific way.
Its mimicry of humans is not really intelligent at all – but is certainly artificial, as in not real!
As an 80-year-old Vietnam War resister and escapee from the U.S., I find AI helpful when I use it to curate factual information untainted by imperial narratives. I mentioned my credentials because after a lifetime of watching the empire’s patterns of behaviour and training my highly sensitive bullshit detector on 60+ years of U.S.-NATO-EU-Five Eyes-Israel-Wall Street-corporate media and every conceivable institution that spews nothing but nonstop lies, I am able to click away from poisonous garbage and not allow it to infiltrate my mind. I feel competent, iow, to use AI because I am pretty sure know when I am being lied to.
A real problem, I think, with AI is that young people, who do not have decades of exposure to analysts debunking lies and decades of smelling lies themselves, are at the mercy of narratives that come at us relentlessly from every direction. I have been living in Canada since 2012, and people — activists — I assumed would know better are completely brainwashed, not only with “Russia!”, “China!”, “Putin!” nonsense, but also with finance sector narratives blaming everyone but itself for 34% of Canadian renters paying more than 50% of their pay on rent.
The ruling class are ruthless. The empire will kill anyone, destroy anything, thoroughly brainwash its own young, and never relent in its obsession to dominate everyone on Earth militarily, economically. politically, and ideologically. AI, which is truly an amazing technology, operates, like us, in this poisonous environment. With care and a working familiarity with a shared reality that I am not sure young people are allowed access to anymore, I am of the opinion it can be a useful tool.
The propaganda war has already been lost at a base level because a bunch of marketers managed to convince the general public that generative models which do nothing resembling independent thought are in fact “AI”. They did it to convince a gullible public that there is something particularly legitimate about these models, when they are in fact simple predictive models trained on massive quantities of data. The very first step to critiquing this junk ought to be recognizing that it is not AI.
Isael and the US government used AI to plan the attack pn Iran, during negotiations, twice.
AI is a wonderful invention. It increases efficiency, expands our capacity, and frees attention for deeper work — yet in the wrong hands it can become a tool for selfish or short?sighted aims. The technology is neutral; its impact depends on the integrity of the one who uses it, and discerning the difference depends on wakefulness.
Let’s stay awake!
Indeed, Annette, it all depends on WHO uses AI. The well read, some university graduates, but it’s all depends on how you were brought up. If you come from a well read parent/grandparents who have provided to their children valid knowledge, then in facing AI you know what it’s spitting out are either wholesale lies or MANUFACTURED inaccuracies.
Yes! What is the agenda behind the blitzkrieg marketing of AI ‘solutions’ across social media? Answer: influence, thought control, social engineering… A few tweaks on the server side, and the “manufactured inaccuracies” can become the received wisdom for many users.
Perhaps even more scary is the potential (or is it already actual?) application of AI technology on the server side to generate profiles of users based on their internet use.
______________is a wonderful invention. It increases efficiency, expands our capacity, and frees attention for deeper work — yet in the wrong hands it can become a tool for selfish or short?sighted aims. The technology is neutral; its impact depends on the integrity of the one who uses it, and discerning the difference depends on wakefulness.
Let’s stay awake!
The problem is that there are so many inventions to fill that blank! We have not demonstrated the capacity to treat our inventiveness with perspective (to stay awake), rather we use our capacity to manifest what we imagine to develop solutions to the problems our inventions create….on and on and on. The wise, which I take your comment to be, have counseled restraint and circumspection for thousands of years.
AI is one big fact lie and must be removed and accompanied with a super big fat fine for lying to the citizenry.
“People will be living in a world that they think they understand and know about, but they’ll unknowingly be viewing only empire-approved information.”
People have always lived in a world that they think they know about…when they do not! We have long lived in a social construct, not in direct relationship with the biophysical reality that supports our lives. Machine intelligence is moving the ‘social’ construct even more quickly out of the slow-form adaptive processes of human and environmental relationships; the equivalent of going from moving the disconnection from deeper reality by the shovel full to moving that disconnection with huge ‘million tons an hour’ earth movers.
There are no guardrails on this process; I suspect that human created laws and regulations will only be added to the data collection and used to further the increasing dominance of these AI tools over the information universe…and thus the political, economic and physical spaces. Where this will leave the human animal with its (our) biological designs and limitations will depend on how relevant and useful we are as the emphasis moves from the value of living things to valuing the operations that control events.
There is only bad news on AI, bad news #1, they have lost their way already. Bad news # 2 they are way ahead of schedule.
Get of to SubStack and find This will Hold for April 28, 2026. titled What do Ted Cruz , John Roberts, Brett Kavanaugh and Amy Coney
Barrett have in common? A Stolen Election. It is a 27-28 minute read and has 88 sources listed throughout the document. Starts with
segment titled INFRASTRUCTURE. I quote the that segment here.
“In November 2000, a network of conservative lawyers flew to Florida land intervened to h alt the counting of democratic ballots,
helping secure the Presidency for George H. Bush by a margin of 537 votes. His brother Jeb Bush was sitting Governor. James Baker
and Ted Cruz assembled the legal team which included, John Roberts, Brett Kavanaugh, and Amy Coney Barrett. Five years
later George H. Bush appointed Alexander Acosta as U.S. Attorney for the Southern District of Florida. Acosta a Kirkland and Ellis alum
former clerk for Supreme Court Justice Samuel Alito, would go on to play a central role in one of the most controversial prosecutions in
history. Two years later after that Acosta signed a non-prosecution agreement with Jeffery Epstein that protected all conspirators
whether named or unnamed – including Donald Trump. The deal was negotiated by by Jay Lefkowitz and KENNETH STAR, Acosta’s
former colleagues at Kirkland and Ellis, and approved by Deputy Attorney General Mark Filip, another former Kirkland partner. In
2018, Donald Trump nominated Brett Kavanaugh to the Supreme Court. As noted, Kavanaugh had previously worked on the Bush v.
Gore legal team as a partner at Kirkland and Ellis Two years later Trump nominated Amy Coney Barrett to the Court. She had also
worked on Bush v. Gore as an associate at Miller, Cassidy, Larroca and Lewin – a firm that later merged with Baker Botts, led by the GOP
power broker, James Baker, who oversaw the recount effort for George H. Bush. By October 2025m when Ghislaine Maxwell’s cert
petition reached the Supreme Court, the bench included three former Bush v. Gore lawyers: Roberts, Kavanaugh, and Barrett – two of
whom had been elevated to their positions by Donald Trump. None recused, and the Court declined to hear the appeal, further
protecting Trump, whether named or unnamed as a co-conspirator. Viewed thru the lens of shielding Trump from legal consequences
tied to decades of criminality – and securing his path in the 2024 elelction – the Supreme Court’s Ruling in Trump v. United States and
Trump v. Anderson began to make sense. INSET Shadow: Bottom line this is a Supreme Court stacked with religious zea;ots, willing to
rig elections toward far right candidates – and to shield the adjudicated felon who put many of them on the bench from accountability.
Still quoting from the first section:
Now lit’s take a look at the systems that continue to deliver victories to Republican despite unpopular policies and out right corruption.”
END Quote
The Article continues on with ES&S.
This information must absolutely get into the main stream of the American Citizenry Consciousness Immediately iF Not Sooner
THANK YOU FOR YOUR ATTENTION TO THIS MATTER
Check out the Youtube channel “Lawrence Wilkerson Updates” (@LawrenceWilkersonUpdates).
Deep fakes of Col. Lawrence Wilkerson.
Wilkerson exposes this in an interview with Nima Alkhorshid on “Dialogue Works” channel:
hxxps://www.youtube.com/watch?v=pqKhinZt3wM (52:25 to 53:05)
Sure enough!
The plot$ thicken.
Goodbye Hollywood propaganda, hello ubiquitous AI.
There’s enough propaganda from “real” sources, never mind introducing AI into the mix.