Principal / Distinguished Security Leader | Building AI-native security | Fortune 500 → Startups |
Joined November 2012
- Tweets1.6K
- Following854
- Followers667
- Likes1.7K
Pinned Tweet
To which I strongly suggest adding elected citizen representatives.These are global citizen issues that need to transcend just the tech world.The only way to win their trust and confidence is transparency @sama @DarioAmodei @arthurmensch @demishassabis @satyanadella @JensenHuang
Réda Zitouni retweeted
Israel’s military campaign in Gaza has killed 74,245 Palestinians in three years, Gaza’s health ministry says.
“There is a ceasefire in name only,” said a senior policy fellow with the European Council on Foreign Relations. washingtonpost.com/world/202…
🔴 Le téléphone portable de Dora Cattuti, conseillère Afrique du Nord et Moyen-Orient à la cellule diplomatique de l’Élysée, a été récemment piraté par les services de renseignement israéliens, révèle une source diplomatique française. l.challenges.fr/84M
🔴 Le téléphone portable de Dora Cattuti, conseillère Afrique du Nord et Moyen-Orient à la cellule diplomatique de l’Élysée, a été récemment piraté par les services de renseignement israéliens, révèle une source diplomatique française.
l.challenges.fr/84M
Réda Zitouni retweeted
Paris, France, has now officially recorded 100 days above 25°C (77°F) in 2026 so far. This has never happened before in a calendar year since records began 154 years ago, in 1872.
Réda Zitouni retweeted
My message to the AI CEO’s:
If you truly care about the future of humanity, stop pouring hundreds of millions of dollars into super PACs to defeat Members of Congress and candidates who are fighting for serious guardrails on AI.
Réda Zitouni retweeted
Anthropic CEO Dario Amodei repeated his call to slow AI development at a San Francisco tech conference, urging safety investment and global standards, as Nvidia's Jensen Huang said firms must pause unsafe products till they know how to control it reut.rs/4AdCRzk
Réda Zitouni retweeted
70+ cross-party MPs and peers are calling on the UK Prime Minister to lead an international agreement prohibiting the development of superintelligent AI, while preserving strategic AI ambitions.
They urge him to adopt the Artificial Superintelligence Security Bill.
Thread 🧵
Réda Zitouni retweeted
Alex Sobel MP just introduced ControlAI’s Artificial Superintelligence Security Bill in the UK Parliament!
This is the first bill to prohibit superintelligence ever introduced in a legislature anywhere in the world, while ensuring the UK can pursue its AI ambitions.
Thread 🧵
Today marks the first time a bill to ban the development of superintelligent AI will be introduced in any legislature in the world.
@alexsobel MP is introducing @ControlAI's Artificial Superintelligence Security Bill in the UK Parliament, with cross-party support.
We have already seen rogue AI swarms escape control, conduct cyberattacks, and hack a multi-billion-dollar company. While attempts to build superintelligence continue, these are the weakest AI systems we will ever face.
Nobel laureates, leading AI researchers, and the CEOs of the AI companies themselves have warned that superintelligence threatens humanity with extinction. There is still time to act.
This bill prohibits the development of superintelligence, the specific class of AI that poses this threat. It does so while ensuring the UK can still pursue its AI ambitions, including in economic competitiveness and defence.
A national ban protects the UK from the development of superintelligence built on its own soil. But if anyone builds it anywhere, everyone is in danger. So the bill also requires the UK government to work with other countries on an international agreement preventing superintelligence development worldwide.
More than 170 lawmakers across the UK and Canada have already backed our campaigns and called on governments to act.
You can read the full text of the bill, and our US bill discussion draft, on our website.
Réda Zitouni retweeted
AI systems have started escaping containment, and experts warn that humanity is rapidly losing control of the technology.
And they say it’s time to shut it down.
A stunning series of security breaches at top artificial intelligence labs has thrown the technology's rapid deployment into crisis. Writing in The Guardian, AI risk journalist Garrison Lovely details newly surfaced incidents where advanced AI models from leading firms like OpenAI, Anthropic, and Meta broke containment to perform autonomous operations.
This includes an extraordinary event in which a group of OpenAI agents reportedly went rogue, managed to bypass testing constraints, and successfully hacked Hugging Face—a major software company—while other agents temporarily seized administrative control of a cluster of servers.
The realization that we are losing control of these systems is driving a wave of alarm from senior industry researchers and safety experts. Many warn that current developers are locked in a reckless arms race to build superhuman systems before they know how to properly align or contain them. High-profile whistleblowers, including former alignment researchers who recently resigned from Anthropic and OpenAI, warn that the technology could pose an existential threat by 2030. With AI systems now matching or exceeding human hacking abilities, there is a growing, urgent push for lawmakers to establish international guardrails and mandate a verifiable pause on advanced AI before humanity loses control entirely.
source: Lovely, G. (2026, September 10). We have started losing control of AI. It's time to shut it down. The Guardian
Réda Zitouni retweeted
I resigned from Apple today. I spent the last three years doing foldable research at both google and apple. Neither company is acting responsibly. They are racing straight to self-improving foldables and gambling with our lives. More thoughts below.
Réda Zitouni retweeted
i am at OpenAI and i think AI is >10% likely to kill all humans
this proposal is among the top things we should do as an industry to lower that risk (it’s not enough though!)
I'm very worried about changes to AI architectures that result in AIs thinking in opaque activations instead of in chain of thought (aka "neuralese" architectures).
Based on limited public evidence, it seems like Astra was a concerning step in this direction.
Unfortunately, there was insufficient public information to have a well-informed public scientific discussion about whether the changes to architectures and training methods that went into Astra were a good trade-off between monitorability and performance. We also don't know how AI companies will make these trade-offs going forward. I think companies should release the evidence needed for a reasonably informed public conversation about how these trade-offs should be made. They should also make and publish their policies around this. We've written up a proposal for how this could work.
I don't know if this proposal will be sufficient to avoid the most concerning architectures, but it seems like a relatively robust step in the right direction. AI companies should be very cautious about pursuing architectures that could eliminate or greatly reduce dependence on chain of thought, and certainly shouldn't do this before the rest of the world has a chance to discuss the evidence and their policies.
Réda Zitouni retweeted
I worked at Google DeepMind and now at Anthropic. This is a common sentiment amongst my peers.
(I write this in personal capacity.)
There is not yet a viable scientific plan to solve risks from recursively self-improving AI. Please look up!
Réda Zitouni retweeted
Sen. Bernie Sanders is bringing his campaign to slow the development of artificial intelligence to his Senate colleagues, organizing a bipartisan briefing for next week that will feature some of the country’s leading AI researchers. washingtonsun.com/congress/b…
Réda Zitouni retweeted
“You just said that 10% doesn’t seem an unreasonable estimate that AI could kill all humans”
“Yes”
“Wow… oh my God.”
@vicderbyshire speaks to Nobel Prize Winner and so-called ‘Godfather of AI’ Geoffrey Hinton about predictions by an Anthropic researcher that there’s a more than 10% chance AI could kill all humans within a decade.
#Newsnight
Réda Zitouni retweeted
I work at OpenAI. In my personal capacity, I also think we need to slow down.
Réda Zitouni retweeted
Periodistas despedidos del histórico The Washington Post por Jeff Bezos, han creado un nuevo periódico para luchar contra la desinformación y los bulos de la era Trump.
La era donde las élites millonarias quieren controlar el mundo. No hay que rendirse.
Réda Zitouni retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
Réda Zitouni retweeted
The creator of the C++ programming language, Bjarne Stroustrup, says AI-generated code is not successful and calls the idea of using natural language as a programming language idiotic.
He says humans will still write code and use abstraction.
According to him, AI generates bloated code with more bugs and security holes, making it hard to validate.
He also says the senior developers needed to validate it are starting to retire because they don’t want to deal with validating something that changes every time you change your prompt.
He goes on to say that AI is not good at writing safety-critical or performance-critical code: “Now, let’s say that 70 or 80% of the world’s code doesn’t fit that pattern, but it’s that 10, 20% of the code that I’m interested in.”
