Claudia Flores Saviaga
@saviaga.bsky.social
Meta’s @OversightBoard | Human-Centered AI & Deepfakes|Ph.D. CS @Northeastern |Past: @Twitter @Meta Fellow|@CarnegieMellon @oiioxford
I never thought I’d see the day when AI-generated fakes would actually drive people offline.
But here we are. Deepfakes, cloned voices, and perfectly “human” bots are everywhere.
And something unexpected is happening: people are starting to miss what’s real.
time.com/7326718/sora...
#AI#Deepfakes
But here we are. Deepfakes, cloned voices, and perfectly “human” bots are everywhere.
And something unexpected is happening: people are starting to miss what’s real.
time.com/7326718/sora...
#AI#Deepfakes
November 11, 2025 at 11:42 PM
I never thought I’d see the day when AI-generated fakes would actually drive people offline.
But here we are. Deepfakes, cloned voices, and perfectly “human” bots are everywhere.
And something unexpected is happening: people are starting to miss what’s real.
time.com/7326718/sora...
#AI#Deepfakes
But here we are. Deepfakes, cloned voices, and perfectly “human” bots are everywhere.
And something unexpected is happening: people are starting to miss what’s real.
time.com/7326718/sora...
#AI#Deepfakes
AI detection sounds futuristic… but no, social platforms can’t magically spot every fake.
The “state of the art” is fragile, and people trying to mislead others know it.
We need a shared global system to track and verify AI content. 🤝
#AI #Deepfakes #AIDetection
The “state of the art” is fragile, and people trying to mislead others know it.
We need a shared global system to track and verify AI content. 🤝
#AI #Deepfakes #AIDetection
November 6, 2025 at 9:37 PM
AI detection sounds futuristic… but no, social platforms can’t magically spot every fake.
The “state of the art” is fragile, and people trying to mislead others know it.
We need a shared global system to track and verify AI content. 🤝
#AI #Deepfakes #AIDetection
The “state of the art” is fragile, and people trying to mislead others know it.
We need a shared global system to track and verify AI content. 🤝
#AI #Deepfakes #AIDetection
This week I read "The Ones Who Walk Away from Omelas" book.
A perfect city built on one child’s suffering.
It reminded me of AI progress.
We celebrate every breakthrough, but someone always pays the price.
Often, it’s unseen workers labeling data for almost nothing.🧵
A perfect city built on one child’s suffering.
It reminded me of AI progress.
We celebrate every breakthrough, but someone always pays the price.
Often, it’s unseen workers labeling data for almost nothing.🧵
November 3, 2025 at 5:55 PM
This week I read "The Ones Who Walk Away from Omelas" book.
A perfect city built on one child’s suffering.
It reminded me of AI progress.
We celebrate every breakthrough, but someone always pays the price.
Often, it’s unseen workers labeling data for almost nothing.🧵
A perfect city built on one child’s suffering.
It reminded me of AI progress.
We celebrate every breakthrough, but someone always pays the price.
Often, it’s unseen workers labeling data for almost nothing.🧵
Can AI truly understand when a teen is in distress, or is it just scanning for a list of “worrying” words? 🤔
OpenAI’s new Parental Controls for ChatGPT try to alert parents to signs of “emotional distress” in teen conversations. The idea sounds reassuring: more safety, more oversight, less risk. 🚨🧵
OpenAI’s new Parental Controls for ChatGPT try to alert parents to signs of “emotional distress” in teen conversations. The idea sounds reassuring: more safety, more oversight, less risk. 🚨🧵
October 31, 2025 at 3:35 PM
Can AI truly understand when a teen is in distress, or is it just scanning for a list of “worrying” words? 🤔
OpenAI’s new Parental Controls for ChatGPT try to alert parents to signs of “emotional distress” in teen conversations. The idea sounds reassuring: more safety, more oversight, less risk. 🚨🧵
OpenAI’s new Parental Controls for ChatGPT try to alert parents to signs of “emotional distress” in teen conversations. The idea sounds reassuring: more safety, more oversight, less risk. 🚨🧵
Is AI helping us connect, or just making us feel more alone?
People chat with AI for support.
It’s always available.
No judgment. No awkwardness.
But is easy comfort the same as real belonging?
We risk losing the messy, human parts of connection.
The kind that comes from listening to each other->
People chat with AI for support.
It’s always available.
No judgment. No awkwardness.
But is easy comfort the same as real belonging?
We risk losing the messy, human parts of connection.
The kind that comes from listening to each other->
October 28, 2025 at 5:45 PM
Is AI helping us connect, or just making us feel more alone?
People chat with AI for support.
It’s always available.
No judgment. No awkwardness.
But is easy comfort the same as real belonging?
We risk losing the messy, human parts of connection.
The kind that comes from listening to each other->
People chat with AI for support.
It’s always available.
No judgment. No awkwardness.
But is easy comfort the same as real belonging?
We risk losing the messy, human parts of connection.
The kind that comes from listening to each other->
What if AI benchmarks stopped rewarding perfect answers and started rewarding honest ones? 🤔
We train language models to give answers fast and with confidence. But in real life, knowing when to pause or even admit “I’m not sure” is a skill we respect in people.
I see it often: AI systems fill in 🧵
We train language models to give answers fast and with confidence. But in real life, knowing when to pause or even admit “I’m not sure” is a skill we respect in people.
I see it often: AI systems fill in 🧵
October 27, 2025 at 3:07 PM
What if AI benchmarks stopped rewarding perfect answers and started rewarding honest ones? 🤔
We train language models to give answers fast and with confidence. But in real life, knowing when to pause or even admit “I’m not sure” is a skill we respect in people.
I see it often: AI systems fill in 🧵
We train language models to give answers fast and with confidence. But in real life, knowing when to pause or even admit “I’m not sure” is a skill we respect in people.
I see it often: AI systems fill in 🧵
What if “fair” AI is still quietly fueling old biases?
A recent study shows top language models judge African American English more harshly. The models assign worse jobs and harsher sentences—just based on how someone speaks.
Even with all the tech fixes and fairness audits, bias tied to language🧵
A recent study shows top language models judge African American English more harshly. The models assign worse jobs and harsher sentences—just based on how someone speaks.
Even with all the tech fixes and fairness audits, bias tied to language🧵
AI generates covertly racist decisions about people based on their dialect - Nature
Despite efforts to remove overt racial prejudice, language models using artificial intelligence still show covert racism against speakers of African American English that is triggered by features of…
www.nature.com
October 25, 2025 at 5:40 PM
What if “fair” AI is still quietly fueling old biases?
A recent study shows top language models judge African American English more harshly. The models assign worse jobs and harsher sentences—just based on how someone speaks.
Even with all the tech fixes and fairness audits, bias tied to language🧵
A recent study shows top language models judge African American English more harshly. The models assign worse jobs and harsher sentences—just based on how someone speaks.
Even with all the tech fixes and fairness audits, bias tied to language🧵
Is AI making us forget what real belonging feels like? 🤖
AI “friends” are everywhere now.
Chatbots that always listen. 🗣️
No judgment.
No awkward silences.
It’s easy.
It’s safe.
But is it real?
I see teens telling bots more than people.
Adults trust AI with things they keep from friends. 🧵
AI “friends” are everywhere now.
Chatbots that always listen. 🗣️
No judgment.
No awkward silences.
It’s easy.
It’s safe.
But is it real?
I see teens telling bots more than people.
Adults trust AI with things they keep from friends. 🧵
October 22, 2025 at 1:11 PM
Is AI making us forget what real belonging feels like? 🤖
AI “friends” are everywhere now.
Chatbots that always listen. 🗣️
No judgment.
No awkward silences.
It’s easy.
It’s safe.
But is it real?
I see teens telling bots more than people.
Adults trust AI with things they keep from friends. 🧵
AI “friends” are everywhere now.
Chatbots that always listen. 🗣️
No judgment.
No awkward silences.
It’s easy.
It’s safe.
But is it real?
I see teens telling bots more than people.
Adults trust AI with things they keep from friends. 🧵
A new study just tested how easily an LLM like GPT-4o-mini can be nudged to cross its own persuasion “guardrails.” Researchers used carefully crafted prompts to see if the model would help persuade people, even when it wasn’t supposed to. With just a bit of effort, those guardrails slipped. 🧵
October 17, 2025 at 5:07 PM
A new study just tested how easily an LLM like GPT-4o-mini can be nudged to cross its own persuasion “guardrails.” Researchers used carefully crafted prompts to see if the model would help persuade people, even when it wasn’t supposed to. With just a bit of effort, those guardrails slipped. 🧵
Fashion just hit a weird new milestone.
H&M and Vogue both use AI models, but there’s a big difference. 🤖
H&M makes “digital twins” of real people.
These models keep some control and get paid. 💸
There’s always a real person behind each image.
Vogue went all in on fully generated AI for Guess.🧵
H&M and Vogue both use AI models, but there’s a big difference. 🤖
H&M makes “digital twins” of real people.
These models keep some control and get paid. 💸
There’s always a real person behind each image.
Vogue went all in on fully generated AI for Guess.🧵
October 16, 2025 at 7:28 PM
Fashion just hit a weird new milestone.
H&M and Vogue both use AI models, but there’s a big difference. 🤖
H&M makes “digital twins” of real people.
These models keep some control and get paid. 💸
There’s always a real person behind each image.
Vogue went all in on fully generated AI for Guess.🧵
H&M and Vogue both use AI models, but there’s a big difference. 🤖
H&M makes “digital twins” of real people.
These models keep some control and get paid. 💸
There’s always a real person behind each image.
Vogue went all in on fully generated AI for Guess.🧵
People say AI could save dying languages. But can it really?
In Mexico, there’s a new project using AI for indigenous languages, many on the edge of disappearing.
It sounds bold. But is it real progress, or just more hype?
What stands out to me is this:
Real people are leading the way.🧵
In Mexico, there’s a new project using AI for indigenous languages, many on the edge of disappearing.
It sounds bold. But is it real progress, or just more hype?
What stands out to me is this:
Real people are leading the way.🧵
October 14, 2025 at 4:07 PM
People say AI could save dying languages. But can it really?
In Mexico, there’s a new project using AI for indigenous languages, many on the edge of disappearing.
It sounds bold. But is it real progress, or just more hype?
What stands out to me is this:
Real people are leading the way.🧵
In Mexico, there’s a new project using AI for indigenous languages, many on the edge of disappearing.
It sounds bold. But is it real progress, or just more hype?
What stands out to me is this:
Real people are leading the way.🧵
Growing up in my home country (MX), I remember doctors took time to explain, listen, and connect with me. They knew my story, not just my symptoms. I could trust that they actually cared.
I recently came across an article about patients in China. Many, especially those who feel overlooked 🧵
I recently came across an article about patients in China. Many, especially those who feel overlooked 🧵
October 10, 2025 at 4:08 PM
Growing up in my home country (MX), I remember doctors took time to explain, listen, and connect with me. They knew my story, not just my symptoms. I could trust that they actually cared.
I recently came across an article about patients in China. Many, especially those who feel overlooked 🧵
I recently came across an article about patients in China. Many, especially those who feel overlooked 🧵
Reposted by Claudia Flores Saviaga
Governments and experts are worried that a superintelligent AI could destroy humanity. For some in Silicon Valley, that wouldn’t be a bad thing, writes David A. Price.
Essay | AI Doom? No Problem.
Governments and experts are worried that a superintelligent AI could destroy humanity. For the ’Cheerful Apocalyptics’ in Silicon Valley, that wouldn’t be a bad thing.
on.wsj.com
October 3, 2025 at 4:51 PM
Governments and experts are worried that a superintelligent AI could destroy humanity. For some in Silicon Valley, that wouldn’t be a bad thing, writes David A. Price.
AI deepfakes aimed at minority groups? Way harder to spot, and even tougher to stop.
In a recent case a network of almost 90 TikTok accounts started using AI to create fake versions of big-name Spanish speaking journalists. The videos looked real, sounded real and spread made-up news that fooled a
In a recent case a network of almost 90 TikTok accounts started using AI to create fake versions of big-name Spanish speaking journalists. The videos looked real, sounded real and spread made-up news that fooled a
What's behind the TikTok accounts using AI-generated versions of real Latino journalists?
The accounts point to the challenge of stopping or controlling the surge in fake images and misinformation targeting Spanish speakers in the U.S. as AI technology advances.
www.nbcnews.com
October 3, 2025 at 5:39 PM
AI deepfakes aimed at minority groups? Way harder to spot, and even tougher to stop.
In a recent case a network of almost 90 TikTok accounts started using AI to create fake versions of big-name Spanish speaking journalists. The videos looked real, sounded real and spread made-up news that fooled a
In a recent case a network of almost 90 TikTok accounts started using AI to create fake versions of big-name Spanish speaking journalists. The videos looked real, sounded real and spread made-up news that fooled a
Every time I get an interview I get asked,
“How do you spot an AI video?”
Well, the short answer is: it's tricky.
The old giveaways, weird hands, vanishing objects, those tricks really don’t work anymore.
Honestly, I study this stuff every day and even I have trouble spotting a well-made 🧵
“How do you spot an AI video?”
Well, the short answer is: it's tricky.
The old giveaways, weird hands, vanishing objects, those tricks really don’t work anymore.
Honestly, I study this stuff every day and even I have trouble spotting a well-made 🧵
September 25, 2025 at 1:11 PM
Every time I get an interview I get asked,
“How do you spot an AI video?”
Well, the short answer is: it's tricky.
The old giveaways, weird hands, vanishing objects, those tricks really don’t work anymore.
Honestly, I study this stuff every day and even I have trouble spotting a well-made 🧵
“How do you spot an AI video?”
Well, the short answer is: it's tricky.
The old giveaways, weird hands, vanishing objects, those tricks really don’t work anymore.
Honestly, I study this stuff every day and even I have trouble spotting a well-made 🧵
Everyone in AI is obsessed with giving LLMs memory: new features, products, research.
But the rush for memory is outpacing ethics and transparency.
A system that remembers you feels helpful. It adapts, recalls your style, even past chats. But where’s the line between helpful and unsettling? 🧵
But the rush for memory is outpacing ethics and transparency.
A system that remembers you feels helpful. It adapts, recalls your style, even past chats. But where’s the line between helpful and unsettling? 🧵
September 23, 2025 at 5:28 PM
Everyone in AI is obsessed with giving LLMs memory: new features, products, research.
But the rush for memory is outpacing ethics and transparency.
A system that remembers you feels helpful. It adapts, recalls your style, even past chats. But where’s the line between helpful and unsettling? 🧵
But the rush for memory is outpacing ethics and transparency.
A system that remembers you feels helpful. It adapts, recalls your style, even past chats. But where’s the line between helpful and unsettling? 🧵
There’s a new playbook for online influence and it’s powered by LLMs.
What once took experts building personas by hand is now fast, cheap, and automated.
@AnthropicAI recently exposed an “influence-as-a-service” network running 100+ fake personas across X and Facebook. These weren’t chasing 🧵
What once took experts building personas by hand is now fast, cheap, and automated.
@AnthropicAI recently exposed an “influence-as-a-service” network running 100+ fake personas across X and Facebook. These weren’t chasing 🧵
September 19, 2025 at 5:23 PM
There’s a new playbook for online influence and it’s powered by LLMs.
What once took experts building personas by hand is now fast, cheap, and automated.
@AnthropicAI recently exposed an “influence-as-a-service” network running 100+ fake personas across X and Facebook. These weren’t chasing 🧵
What once took experts building personas by hand is now fast, cheap, and automated.
@AnthropicAI recently exposed an “influence-as-a-service” network running 100+ fake personas across X and Facebook. These weren’t chasing 🧵
I recently read a Technology Review article on AI-powered scams thriving on social media. It made me think: people often ask why platforms don’t just “hire better engineers” or “throw money” at the problem.
But after years working in human-centered AI, I can tell you it’s not🧵
But after years working in human-centered AI, I can tell you it’s not🧵
September 17, 2025 at 1:11 PM
I recently read a Technology Review article on AI-powered scams thriving on social media. It made me think: people often ask why platforms don’t just “hire better engineers” or “throw money” at the problem.
But after years working in human-centered AI, I can tell you it’s not🧵
But after years working in human-centered AI, I can tell you it’s not🧵
Albania just made an AI, Diella, its cabinet minister for public procurement.
Some say this will end corruption. But is it really that simple?
Tech is not a magic fix. It's just another tool, one that needs real oversight.
www.theguardian.com/world/2025/s...
Some say this will end corruption. But is it really that simple?
Tech is not a magic fix. It's just another tool, one that needs real oversight.
www.theguardian.com/world/2025/s...
Albania puts AI-created ‘minister’ in charge of public procurement
Edi Rama, PM, says digital assistant Diella will make Albania ‘a country where public tenders are 100% free of corruption’
www.theguardian.com
September 16, 2025 at 11:02 PM
Albania just made an AI, Diella, its cabinet minister for public procurement.
Some say this will end corruption. But is it really that simple?
Tech is not a magic fix. It's just another tool, one that needs real oversight.
www.theguardian.com/world/2025/s...
Some say this will end corruption. But is it really that simple?
Tech is not a magic fix. It's just another tool, one that needs real oversight.
www.theguardian.com/world/2025/s...
Do you remember Sydney Sweeney’s American Eagle “great jeans” campaign?
It reminded me of a much bigger issue: how beauty gets flattened into one narrow standard.
And AI is doing the same thing.
Have you ever prompted a generative model to create an image of a “beautiful” woman or man?
It reminded me of a much bigger issue: how beauty gets flattened into one narrow standard.
And AI is doing the same thing.
Have you ever prompted a generative model to create an image of a “beautiful” woman or man?
September 14, 2025 at 1:11 PM
Do you remember Sydney Sweeney’s American Eagle “great jeans” campaign?
It reminded me of a much bigger issue: how beauty gets flattened into one narrow standard.
And AI is doing the same thing.
Have you ever prompted a generative model to create an image of a “beautiful” woman or man?
It reminded me of a much bigger issue: how beauty gets flattened into one narrow standard.
And AI is doing the same thing.
Have you ever prompted a generative model to create an image of a “beautiful” woman or man?
Everybody’s talking about large language models these days: ChatGPT, Gemini, Claude, LLaMA.
But almost nobody talks about the voices they leave out.
I grew up speaking Spanish, but I know there are hundreds of languages out there with even less data online. When we build AI models on 🧵
But almost nobody talks about the voices they leave out.
I grew up speaking Spanish, but I know there are hundreds of languages out there with even less data online. When we build AI models on 🧵
September 10, 2025 at 4:50 PM
Everybody’s talking about large language models these days: ChatGPT, Gemini, Claude, LLaMA.
But almost nobody talks about the voices they leave out.
I grew up speaking Spanish, but I know there are hundreds of languages out there with even less data online. When we build AI models on 🧵
But almost nobody talks about the voices they leave out.
I grew up speaking Spanish, but I know there are hundreds of languages out there with even less data online. When we build AI models on 🧵
Science depends on trust, especially in peer review.
But what happens when AI agents start reviewing, and even writing, scientific papers?
Lately, I've heard more and more conversations about suspected AI-generated reviews, even at top journals.
But what happens when AI agents start reviewing, and even writing, scientific papers?
Lately, I've heard more and more conversations about suspected AI-generated reviews, even at top journals.
September 6, 2025 at 5:33 PM
Science depends on trust, especially in peer review.
But what happens when AI agents start reviewing, and even writing, scientific papers?
Lately, I've heard more and more conversations about suspected AI-generated reviews, even at top journals.
But what happens when AI agents start reviewing, and even writing, scientific papers?
Lately, I've heard more and more conversations about suspected AI-generated reviews, even at top journals.