mahsa alimardani
@mahsaalimardani.bsky.social
Technology and human rights. Associate Director of Technology Threats and Opportunities at @witnessorg.bsky.social | Doctoral Researcher @oii.ox.ac.uk
1/New with Afsaneh Rigot of @thedecenter.bsky.social with work from @witnessorg.bsky.social for @carnegieendowment.org: "Generative AI and Atrocity Denial in War." How AI has turned doubt itself into a weapon, from Iran to Gaza to Sudan. Thread: carnegieendowment.org/research/202...
Generative AI and Atrocity Denial in War
In the fog of modern war, generative AI offers every side a simple transaction: Trade documented reality for permanent doubt.
carnegieendowment.org
In @techpolicypress.bsky.social shirin anlen and I on a foreseeable fiasco: how Google briefly turned 1 of world's most trusted sources of evidence into a generator of plausible fakes, and why tech cos must consult civil society experts before launch, not after www.techpolicy.press/google-earth...
Google Earth AI Fiasco Underscores Why Tech Firms Must Listen to Outside Experts
It's just one more parable in a long line of product releases by AI firms with potentially dangerous consequences, write shirin anlen and Sam Gregory.
techpolicy.press
A personal essay on grief, memory and Iran but most important about my baba. time.com/article/2026...
The Far Gates: A father, a daughter, and flights from Iran
"My father left Iran carrying the hope of one day being buried there. He never made it back, but grief spoken into the open is the closest thing to a homeland that exile will ever leave you," writes M...
time.com
Does Europe have what it takes for AI sovereignty? 👀 @nytimes hosted one of the most charged debates at #VivaTech 2026 and the audience had the final say👇
@mahsaalimardani.bsky.social, Associate Director, Technology Threats and Opportunities at @witnessorg.bsky.social reads an extract from her essay ‘Layered trust signals, not silver bullets: what Iran shows the UK about AI and public trust’ which is featured in the 2026 Full Fact Report.
NEW from me: Grok has continued to host sexualized and demeaning images of celebrities, a review of public posts from Grok Imagine has found. As well as this, people have been creating full nudity (against Grok’s terms) and sex scenes with the Musk-owned AI
Grok Is Still Hosting Sexualized Deepfakes of Famous Women
A WIRED investigation found dozens of “nudified” deepfake images and videos on Grok's website, including nonconsensual depictions of celebrities and at least one prominent US politician.
wired.com
This is a very interesting and unfortunately topical conversation between @mahsaalimardani.bsky.social and @mjgault.bsky.social on how belligerents in the US/Israeli war on Iran are using GenAI and other information manipulation to sway public opinion: podcasts.apple.com/lu/podcast/a...
Navigating Reality and War During the Age of AI Propaganda
Podcast Episode · Angry Planet · 5 June · 59min
podcasts.apple.com
Join us tmrw. to discuss how actors engaged in protection efforts are stepping up to the challenge of addressing digital risks in conflict and crisis. Register: www.accessnow.org/event/digita... @mahsaalimardani.bsky.social @witnessorg.bsky.social @cloudflare.social @oxfaminternational.bsky.social
Access Now@accessnow.org · 3mo agoTechnology can amplify risks for civil society in the context of war and conflict — but centering digital rights can also enhance humanitarian and protection responses. Join our Digital Security Helpline’s webinar on May 19 to learn more 🔽 www.accessnow.org/event/digita...
Join us ONLINE (May 19th-3pm CEST) to explore digital threats in conflict with @mahsaalimardani.bsky.social ( @witnessorg.bsky.social ), Jocelyn Woolbright ( @cloudflare.social ), Luis Eduardo BARRUETO ( @accessnow.org Digital Security Helpline), and Mia Marzotto ( @oxfaminternational.bsky.social )
Access Now@accessnow.org · 3mo agoTechnology can amplify risks for civil society in the context of war and conflict — but centering digital rights can also enhance humanitarian and protection responses. Join our Digital Security Helpline’s webinar on May 19 to learn more 🔽 www.accessnow.org/event/digita...
Yesterday, the Paris prosecutor's office announced an escalation of its investigation of Elon Musk and his social network X, seeking formal charges on a range of claims, including complicity in the possession of child sexual abuse material. We updated our timeline with the latest.
Tracking the Paris Prosecutor's Investigation into Elon Musk's X
A raid on the offices of X in Paris followed a year-long investigation that began in January 2025.
techpolicy.press
Mahsa's article on the eight Iranian protestors that Trump made appear synthetic is a must read
The Real Iranian Women Protesters Trump Made Look Synthetic
Political prisoners are packaged into AI-enhanced collages and unverified execution claims. The Iranian state is the only beneficiary, writes Mahsa Alimardani.
techpolicy.press
really great conversation with @mahsaalimardani.bsky.social about what AI-powered propaganda is doing to the information space
Navigating Reality and War During the Age of AI Propaganda
On the morning of April 21, Trump posted an image of eight women on Truth Social, claimed they were Iranian dissidents set to be executed, and demanded that Tehran release them.
angryplanetpod.com
France has summoned Musk and Yaccarino to face preliminary criminal charges over Grok-generated CSAM and sexually explicit deepfakes. The Durov/Telegram case was about failing to moderate user content.
🧵My paper on "Uneven Datafication" just came out! It draws on literature on coloniality, uneven development and dependency. Uneven datafication operates through three interrelated dynamics: journals.sagepub.com/doi/10.1177/...
New on @indicator.media: At 11 a.m. on Wednesday morning, I discovered that my partner Mia has a hidden Tinder Platinum account. Except Mia is not actually my partner. She isn’t even a real person. The image I uploaded was created with Google AI. So, it turns out, was the Tinder profile.
How "Cheater Buster" websites use fake info and dubious claims to trick people into paying
Indicator found an ecosystem rife with spurious claims, fake products, and AI-generated matches
indicator.media
New analysis by Valerie Wirtschafter (@vwirtschafter.bsky.social): AI-generated content flagged on X has surged to record highs since the Iran conflict began. The data point to a coordinated effort to use AI as a tool of information warfare—eroding public support and sowing confusion at scale.
Generative AI as a weapon of war in Iran | Brookings
The Iran conflict demonstrates the concrete value of content moderation. Such moderation should flag source reliability in crisis conditions.
brook.gs
Will be part of this @accessnow.org webinar with @giuliowolfe.bsky.social and and others on May 19. Join us! www.accessnow.org/event/digita...
Digital security in war and conflict: challenges for civil society and tools for resilience - Access Now
Join the next webinar organized by the Digital Security Helpline, to discuss key trends and strategies to keep at-risk actors safe online.
accessnow.org
Access Now@accessnow.org · 3mo agoTechnology can amplify risks for civil society in the context of war and conflict — but centering digital rights can also enhance humanitarian and protection responses. Join our Digital Security Helpline’s webinar on May 19 to learn more 🔽 www.accessnow.org/event/digita...
Day 66 of Iran's internet shutdown, the longest nationwide blackout. Over 90 million people cut off. Tomorrow I'm joining @farzdusa.bsky.social and J. Alex Dalessio at SplinterCon Online on direct-to-cell, satellite internet, and overcoming shutdowns. Register: splintercon.net/splintercon-...
Reuters Institute. NEW on our site How AI is thickening the fog of war With the US and Iran using AI and memes to obstruct conflict coverage, our own Gretel Kahn spoke to @mantzarlis.com @mahsaalimardani.bsky.social Emerson Booking and Sam Dubberley from @hrw.org
Reuters Institute@reutersinstitute.bsky.social · 3mo agoNEW on our site How AI is thickening the fog of war With the US and Iran using AI and memes to obstruct conflict coverage, our own Gretel Kahn spoke to @mantzarlis.com @mahsaalimardani.bsky.social Emerson Booking and Sam Dubberley from @hrw.org reutersinstitute.politics.ox.ac.uk/news/trollin...
NEW on our site How AI is thickening the fog of war With the US and Iran using AI and memes to obstruct conflict coverage, our own Gretel Kahn spoke to @mantzarlis.com @mahsaalimardani.bsky.social Emerson Booking and Sam Dubberley from @hrw.org reutersinstitute.politics.ox.ac.uk/news/trollin...
Trolling, memes and deepfakes: How AI is thickening the fog of war
The US and Iran are using technology to obstruct coverage of the current conflict. Four experts discuss how AI is reshaping war reporting and the news ecosystem.
reutersinstitute.politics.ox.ac.uk
This was the most devastating part of the Rights Con debacle. The loss in not having the opportunity to attend, learn and further our movement through what was going to be one of the most thoughtful and productive convenings broke my heart. Please donate to @thedecenter.bsky.social if you can.
De|Center@thedecenter.bsky.social · 3mo agoOur Design from the Margins workshop was cancelled after the Zambian government shutdown RightsCon. We moved fast because the safety of our community came first. It always will. Our loss was huge, but the work continues. See our full statement. the-decenter.ghost.io/update-the-f...
Our Design from the Margins workshop was cancelled after the Zambian government shutdown RightsCon. We moved fast because the safety of our community came first. It always will. Our loss was huge, but the work continues. See our full statement. the-decenter.ghost.io/update-the-f...
Update: The Forced Cancellation of the Design from the Margins Workshop in Lusaka
💡The cancellation of RightsCon – which was to bring 5,000 people together – mere days before it was set to begin, is itself a serious concern for the digital rights community. As we could not determin...
the-decenter.ghost.io
I wrote about RightsCon's cancellation after Zambia's government announced it would postpone the event to sort out its concerns over "thematic issues." With thousands about to depart (or already en route) to Lusaka, it's a huge blow to the digital and human rights field.
RightsCon Canceled After Zambia Requires ‘Full Alignment’ With 'National Values'
Organizers canceled the event just days before thousands were set to descend on Lusaka to discuss issues at the intersection of technology and human rights.
techpolicy.press
I had missed RightsCon for a few years and was genuinely looking forward to being in a room with this community again. The world is bleak. We find light in each other. Watching a government try to snuff that out is a serious blow, and it should be treated as one.
RightsCon@rightscon.org · 3mo agoIt is with heavy hearts that we share: RightsCon will not proceed in Zambia or online.
Misunderstanding of our AI reality as a binary AI vs. Real is furthering eroding the truth in the information space, best exemplified by this week's Trump vs. Islamic Republic disinformation tit for tat over real Iranian women protesters. @witnessorg.bsky.social @techpolicypress.bsky.social
Tech Policy Press@techpolicypress.bsky.social · 3mo agoAuthentic Iranian political prisoners are being packaged into AI-enhanced collages and unverified execution claims and fed into a propaganda noise machine. The Iranian state is the only beneficiary, writes Mahsa Alimardani.
Well, the images are AI-modified/generated, the women are real (verifiably six of them, at least), only one of them is slated to be executed, and Tehran claims no concessions have been made www.theverge.com/policy/91718...
The Iranian women Trump ‘saved’ from execution are simultaneously real and AI-manipulated
Liars are lying again
theverge.com
Fiona "Fi" Webster 🌎🌍🌏@fiona-webster22.bsky.social · 3mo ago47 was mercilessly mocked when he reposted the claim Iran was going to execute eight women because the post was fake with AI-generated images of those women. Now he's doubled-down and claims Iran cancelled the hangings. What a fucking buffoon.
I published this yesterday. www.eff.org/deeplinks/20... Today I learned that my friend Ahmed Shihab Eldin is among those detained. www.theguardian.com/world/2026/a...
War as a Pretext: Gulf States Are Tightening the Screws on Speech—Again
War does not only reshape borders. It also reshapes what can be seen, said, and remembered. When governments invoke “misinformation” during wartime, they often mean something simpler: speech they do n...
eff.org
“Those who stared too long into the Lego abyss came out sounding completely deranged.” @sarahjeong.bsky.social continues to be one of my favourite writers and honoured my scraps of thought were featured in this alongside @azadehakbari.bsky.social and @thedecenter.bsky.social’s Afsaneh Rigot.
How Iran out-shitposted the White House
Tehran’s meme warfare became an uncanny mirror of American governance
theverge.com
Quoted in @apnews.com on Iran's AI meme war. The tell: if you can generate and upload this content at volume while Iran's internet is near-totally shut down, you are officially or unofficially cooperating with the regime. Bandwidth is evidence. apnews.com/article/ai-m...
Pro-Iran groups have used AI to troll Trump and try to control the war narrative
Pro-Iran groups have used artificial intelligence to create slick internet memes in English to try to shape the narrative during the war against the U.S. and Israel, and foster opposition to it.
apnews.com
