Sam Gregory

All the rape, torture, and murder will be brushed under the carpet, a great lesson for any future war criminals.

How Gemini 3 transforms video into structured, searchable datasets: Implications for human rights, and digital forensics

www.linkedin.com/pulse/how-ge... (LinkedIn)

sanjanah.wordpress.com/2025/11/19/h... (WordPress)

@samgregory.bsky.social

The Maybe

Great @themaybe.org podcast with the inimitable @alixdunn.com diving into AI, video, trust and human rights. Loved deep diving into how what we've learned at @witnessorg.bsky.social over decades of video activism drives our work to defend visual truth in age of Sora www.themaybe.org/podcast/stra...

#mustread Every day I wish there was a way to keep all the AI generated content out of my IG feed. So tired of seeing 'people' who don't actually exist.

“the weaponization of doubt”

YouTube video by All Tech Is Human When 'Seeing is Believing' Falters: AI, Visual Truth and Epistemic Fracture | Sam Gregory

“the deeper damage is a fog of doubt settling over everything we see”

@samgregory.bsky.social urges us to consider that soon our shared basis of reality will no longer be trustworthy and risks being undermined by generative video... and offers how to fight that.
www.youtube.com/watch?v=VOBA...

The Body as a Password: The Cost of Face ID Convenience is costing people their freedom. Research shows that features like Face ID and fingerprints enable violent, forced phone access by police and authorities, writes Afsaneh Rigot.

"Research shows that features like Face ID and fingerprints enable violent, forced phone access by police and authorities" the-decenter.ghost.io/the-body-as-... by Afsaneh Rigot

Thanks @raustinjr.bsky.social!

@samgregory.bsky.social has written a fantastic piece on the risks of deep fakes and what can be done about them. As he correctly points out: "When any piece of evidence can be credibly dismissed as synthetic and any fabrication defended as real, we lose the ability to establish shared truth."

Sora Is Here. The window to save visual truth is closing Opinion: Sam Gregory argues that generative video is undermining the notion of a shared reality, and that we need to act before it’s lost forever

"How we choose to act now will determine whether our shared basis for reality — what we see and hear — remains trustworthy."

Generative video's threat to visual truth and what we can do about it, and why concerted action matters now, in @transformernews.ai

www.transformernews.ai/p/sora-is-he...

Sora Is Here. The window to save visual truth is closing Opinion: Sam Gregory argues that generative video is undermining the notion of a shared reality, and that we need to act before it’s lost forever

"How we choose to act now will determine whether our shared basis for reality — what we see and hear — remains trustworthy." Witness's @samgregory.bsky.social on Sora's threat to visual truth:

Call to action here from @samgregory.bsky.social that it would be nice if OpenAI and others paid attention to.

Last week I wrote about the racist narratives being pushed about SNAP benefits, including unhinged AI videos on TikTok

Fox News however ran a story claiming those videos were real until they realised their mistake and changed the story without admitting their mistake.

www.wired.com/story/disinf...

Thanks for the reminder @b-cavello.bsky.social! Was working to fit character limits and wasn't aware of this 🤦

YouTube video by All Tech Is Human When 'Seeing is Believing' Falters: AI, Visual Truth and Epistemic Fracture | Sam Gregory

𝗪𝗵𝗲𝗻 '𝗦𝗲𝗲𝗶𝗻𝗴 𝗶𝘀 𝗕𝗲𝗹𝗶𝗲𝘃𝗶𝗻𝗴' 𝗙𝗮𝗹𝘁𝗲𝗿𝘀: 𝗔𝗜, 𝗩𝗶𝘀𝘂𝗮𝗹 𝗧𝗿𝘂𝘁𝗵 𝗮𝗻𝗱 𝗘𝗽𝗶𝘀𝘁𝗲𝗺𝗶𝗰 𝗙𝗿𝗮𝗰𝘁𝘂𝗿𝗲:I recently spoke about how AI-generated media-like OpenAI's Sora app—is fracturing our shared sense of visual reality, how we're running out of time to address it, and 4 urgent actions to take
www.youtube.com/watch?v=VOBA...

Sora app's hyperreal AI videos ignite online trust crisis as downloads surge Scrolling through the Sora app can feel a bit like entering a real-life multiverse.

"𝑇ℎ𝑒 𝑏𝑖𝑔𝑔𝑒𝑠𝑡 𝑟𝑖𝑠𝑘 𝑤𝑖𝑡ℎ 𝑆𝑜𝑟𝑎 𝑖𝑠 𝑡ℎ𝑎𝑡 𝑖𝑡 𝑚𝑎𝑘𝑒𝑠 𝑝𝑙𝑎𝑢𝑠𝑖𝑏𝑙𝑒 𝑑𝑒𝑛𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑖𝑚𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 𝑡𝑜 𝑜𝑣𝑒𝑟𝑐𝑜𝑚𝑒, 𝑎𝑛𝑑 𝑡ℎ𝑎𝑡 𝑖𝑡 𝑒𝑟𝑜𝑑𝑒𝑠 𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑐𝑒 𝑖𝑛 𝑜𝑢𝑟 𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑡𝑜 𝑑𝑖𝑠𝑐𝑒𝑟𝑛 𝑎𝑢𝑡ℎ𝑒𝑛𝑡𝑖𝑐 𝑓𝑟𝑜𝑚 𝑠𝑦𝑛𝑡ℎ𝑒𝑡𝑖𝑐. 𝐼𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑓𝑎𝑘𝑒𝑠 𝑚𝑎𝑡𝑡𝑒𝑟, 𝑏𝑢𝑡 𝑡ℎ𝑒 𝑟𝑒𝑎𝑙 𝑑𝑎𝑚𝑎𝑔𝑒 𝑖𝑠 𝑎 𝑓𝑜𝑔 𝑜𝑓 𝑑𝑜𝑢𝑏𝑡 𝑠𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑜𝑣𝑒𝑟 𝑒𝑣𝑒𝑟𝑦𝑡ℎ𝑖𝑛𝑔 𝑤𝑒 𝑠𝑒𝑒”
www.latimes.com/business/sto...

How did we beat Nigel Farage and Reform in Caerphilly? We stood by our convictions | Rhun ap Iorwerth Welsh voters have rejected Labour and Starmerite managerialism. By offering hope and not division we showed the populist right can be beaten, says Rhun ap Iorwerth, leader of Plaid Cymru

www.theguardian.com/commentisfre...

1/Citizen Lab’s new report on PRISONBREAK is a must-read. It exposes an AI-enabled influence operation targeting Iran with deceptive media, from the infamous Evin Prison gates video to deepfake music.

A prominent US family has been cashing in on DRC carbon credits while local communities decry unmet promises Since 2004, a prominent US family has made millions through the sale of carbon credits in the DRC. Meanwhile, the community who lives there has been left in poverty.

globalvoices.org/2025/09/10/a...

10 things to know about the camera on Pixel 10 Here are just some of the things you can do with the camera on the new Pixel 10 smartphone lineup.

Pixel 10 out - 1st smartphone integrating C2PA in native camera app, shows content credentials. Reasons impt:
*Reality matters: key to show authenticity of photos
*Shows 'how' AI and human mixed in creation + edits
*Focuses on recipe: on the how, not the who (privacy!)
blog.google/products/pix...

Starmer v Starmer: why is the former human rights lawyer so cautious about defending human rights? The long read: Given his career as a human rights barrister, it was hoped Keir Starmer would be the PM to restore the UK’s commitment to international law. The reality tells a different story

Good morning, I've written a piece about the prime minister. I spoke to several dozen sources, including government insiders, former legal colleagues and campaigners to ask: what happened to Keir Starmer the heroic human rights lawyer? www.theguardian.com/politics/202...

4/ Same fundamentals of consent, disclosure, and hopefully integrity to the person's identity/speech. But that integrity is clouded by chatbot hallucinations, by a bot's frequent trained instinct 'to please,' and also most importantly by filling in blanks from past context to present-day advocacy

3/ How does an interactive chatbot-driven avatar differ from static resurrection deepfakes such as murdered Mexican journalist Javier Valdez calling for justice, a victim impact statement from someone dead in Arizona, or President Suharto coming back from beyond the grave campaigning for his party?

2/ Jim Acosta "interviewed" a representation of one of the victims of the Parkland shootings, Joaquin Oliver, created by his father. It raised key questions about what it means to bring a griefbot - an interactive chatbot that relatives and friends interact with - into the public sphere.

I made an AI clone of my dead son - and let a journalist interview him An interview between a US journalist and an AI clone of a teenager who was killed during shooting at his school has prompted criticism online.

Griefbots are increasingly common in personal usage while static resurrection deepfakes are used for advocacy and entertainment, but what does it mean to interact with an avatar of a dead person for political advocacy? I talked to BBC about THAT interview by Jim Acosta.
www.bbc.com/news/videos/...

4/ Same fundamentals of consent, disclosure, and hopefully integrity to the person's identity/speech. But that integrity is clouded by chatbot hallucinations, by a bot's frequent trained instinct 'to please,' and also by the nature of filling in the blanks from past context to present-day advocacy

3/ How does an interactive chatbot-driven avatar differ from static resurrection deepfakes such as the murdered Mexican journalist Javier Valdez, a victim impact statement from someone dead in Arizona, or a President Suharto coming back from beyond the grave? All usages over the past years.

2/ Jim Acosta "interviewed" a representation of one of the victims of the Parkland shootings, Joaquin Oliver. It raised key questions about what it means to bring a griefbot - an interactive chatbot that relatives and friends interact with - into the public sphere.

Iran-Israel AI War Propaganda Is a Warning to the World The war signaled critical challenges in the growing gap between the sophistication of AI deception and analysts’ ability to detect it, heralding a frightening new phase in global conflict.

"AI-generated content related to the Iran-Israel conflict has taken disinformation to an industrial level,"

@mahsaalimardani.bsky.social and @samgregory.bsky.social write for @carnegieendowment.org on the rise of AI war propaganda.