Mark Stockley (@MarkStockley) header
Mark Stockley (@MarkStockley) profile picture

Host of the award-winning AI Fix podcast and cybersecurity evangelist for @Threat_Down.

Oxford
Joined November 2007
Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)Photo by Mark Stockley (@MarkStockley)
Mark Stockley retweeted
AI is finding security flaws faster than companies can patch them. Join @CoreyNoles and @Threat_Down's @MarkStockley LIVE as we break down the AI patch tsunami, @OpenAI's 948 Linux findings, and what comes next for cybersecurity. 🔴 Watch LIVE youtube.com/live/fJz57zR8nLk…
1
2
4
596
Mark Stockley retweeted
Criminal frontier AI was supposed to be prohibitively expensive. Our latest research cited in @thecubetv found access starts at $12.99 a month. Kriminal isn't a custom-built model. It's Grok in a trench coat. The proof is in its own code. 🧥👀 bit.ly/4x8Oi9J
1
1
1
651
Mark Stockley retweeted
This OAI/HF piece from @GaryMarcus and Zack Korman -- whose main takeaway is that AI labs should practice defense in depth and suffer repercussions when they don't -- is representative of the response from many security folks. This position is not even wrong, it's just that it repeats the obviously true security catechism of the past 20 years in the face of a milestone event that marks the beginning of a new era in our field. It's as though we've discovered Winter is Coming and we're mainly focused on the banal politics of who's to blame in how we found out. To be clear: should OAI/Anthropic/Irregular/Meta/etc. do the things we've been saying forever and pay a price when they don't? Yes. Is it dangerous that they're not doing them? Yes. But also -- thought leadership should engage the new questions -- like: - What do we do about the fact that your median regional hospital network has *worse* security than OpenAI, Anthropic, Huggingface, and Meta? How do we secure tens of thousands of critical organizations quickly? What's AI's role here and what isn't AI's role here? - What are the potentially destabilizing geopolitical implications of hacking agents that can turn amateur-level non-state cyber actors into elite state-level actors now? - What will cybercrime look like when amateurish ransomware affiliates suddenly have the capabilities of elite nation states and what do we do about this? - To what extent will we need to guardrail our networks from internal loss of control incidents now given the business pressures to turn more and more engineering functions over to agents?
Now live (and free) at Marcus on AI: takeways on the OpenAI / Hugging Face attack, coauthored with @ZackKorman, focusing not so much on what the AI did as on what OpenAI should have done. garymarcus.substack.com/p/5-…
8
27
2
138
12,896
Mark Stockley retweeted
24,650 servers were found leaking password data through a 22-year-old protocol flaw baked into their motherboards. 🤯 🖥️ Our own @MarkStockley talked to @HackRead about why there's no single patch for this one. 👉 bit.ly/3TsEUPr
1
1
703
Mark Stockley retweeted
🎙️ @MarkStockley and Fran Bourke joined the Kite Channel Lounge podcast to talk AI-powered threats and why "set it and forget it" security doesn't cut it anymore. Thanks to our friends at Kite for having us, this was a great convo! ▶️ bit.ly/44LisDr
1
3
451
Mark Stockley retweeted
thank you for having me on the podcast @MarkStockley this was a lot of fun! We discussed: + Hassabis proposes the Einstein test for AGI. + An AI agent went rogue on a safety researcher's inbox. + Anthropic accuses Chinese AI labs of distillation attacks on Claude podcasts.apple.com/gb/podcas…
2
3
436
Mark Stockley retweeted
⚠️ Never use an LLM to generate a password. ⚠️ A password is only safe if it’s hard to guess. That means no predictable patterns, no 123456, no p@ssword!, no dictionary words, no pet names, no favourite sports teams. It needs to be random. Something like: G7$kL9#mQ2&xP4!w
1
1
134
Mark Stockley retweeted
When it comes to AI you can have safe or honest, not both. Research suggests that the neurons stopping hallucinations are the same neurons stopping the AI from telling you how to make ricin.
1
1
81
Mark Stockley retweeted
What happens when an AI stops chatting and starts acting? 🦾 @CoreyNoles joins @TheAIFix to unpack @openclaw, an AI agent that is less chatbot and more autonomous operator with hands and spending power. This episode is not about better answers. It is about AI that can take action in the real world. Listen to The AI Fix: 📱Spotify open.spotify.com/episode/2cf… 🎙️Apple Podcasts podcasts.apple.com/gb/podcas… #OpenClaw #AIAgents #AutonomousAgents #GenerativeAI #AIInnovation #Automation #AGI #EmergingTechnology
1
1
1
2
380
Mark Stockley retweeted
Do you wish your AI was a bit less... demonic? A little less ghostly and a bit more consultant-y? Strangely, even if you don't, you probably do.
1
1
114
Mark Stockley retweeted
#AI is transforming the threat landscape, and the pace of change is accelerating. In this FREE webinar, sponsored by @Threat_Down, @MarkStockley will explore how AI is changing #cybersecurity and how to protect your business. Sign up: securitytoday.com/webcasts/2…
2
1
113
Mark Stockley retweeted
BREAKING: Incredible ChatGPT Plus Limit Updates
105
109
20
4,062
340,697
The methods we used to trace the thoughts of Claude are now open to the public! Today, we are releasing a library which lets anyone generate graphs which show the internal reasoning steps a model used to arrive at an answer.
Our interpretability team recently released research that traced the thoughts of a large language model. Now we’re open-sourcing the method. Researchers can generate “attribution graphs” like those in our study, and explore them interactively.
17
162
20
1,828
181,832
Mark Stockley retweeted
Claude Sonnet 4 on ARC-AGI Semi Private Eval Base * ARC-AGI-1: 23%, $0.08/task * ARC-AGI-2: 1.2%, $0.12/task Thinking 16K * ARC-AGI-1: 40%, $0.36/task * ARC-AGI-2: 5.9%, $0.48/task Sonnet 4 sets new SOTA (5.9%) on ARC-AGI-2
39
67
15
569
96,030
Mark Stockley retweeted
The Cl0p ransomware gang is well and truly back—it listed a staggering 335 victims on its leak site in February. The extraordinary haul is a result of the group’s unique tactics.
1
2
2
600
Mark Stockley retweeted
these ARC-AGI tasks are getting out of hand
139
626
253
26,877
11,727,579
We have just completed the last trip to the Great Pacific Garbage Patch for 2024, resulting in our total catch of over 500,000 kg (1.1m lb) from the GPGP so far. Here are some facts about our ocean operations to date:
73
557
43
2,352
100,703
Mark Stockley retweeted
In September, there were 370 known ransomware attacks, with a 63% increase in activity from ‘dark horse’ ransomware groups compared to last year. These smaller, less prominent gangs (what we're calling 'dark horse' gangs) now account for 30% of all known attacks, up from 20%.
1
2
3
653
Mark Stockley retweeted
A new episode of The AI Fix will be out VERY SOON! In the meantime, enjoy this restored version of @WilliamShatner performing Rocket Man, um, for no reason 😉 youtu.be/8wI4jMxveyI?feature…
1
2
137