Blake Thompson Heuer

Security PR @Cisco. @Synack, @POLITICO / @EENewsUpdates alum. "Cyber" prefix aficionado.

[Cisco AI @CiscoAI]

Frontier attacker reasoning, meet defender-grade controls.

Cisco AI Defense is among the first to bring @AnthropicAI's Claude Mythos to AI red teaming, helping security teams uncover risks unique to their AI agents before adversaries do.

Every test authorized, attributed, and audited.

Learn more: https://t.co/Iquf3uYSA2

[Blake Thompson Heuer @BlakeOnSecurity]

@miriamwldvgl wait, no Harwood, ND, dateline? Way to work the phones!

[Blake Thompson Heuer @BlakeOnSecurity]

@sammsabin

[WIRED @WIRED]

Named “Volta-Autodoxer,” the recently uncovered malware uses AI tools to collect troves of data about victims, from login credentials and credit card numbers to Discord sessions and UberEats receipts. https://t.co/Ivma1ibkWn

[Sam Sabin @sammsabin]

Cybersecurity Awareness Month actually meaning something this year is the ultimate Uno reverse

[Blake Thompson Heuer @BlakeOnSecurity]

The PHP vulns are a CEO-level issue.

[Chuck Robbins @ChuckRobbins]

Great work by the @TalosSecurity team getting this out quickly! If anything in your business runs on PHP, this is critical info and requires every company to move quickly. The old habit of baking patches into the next general purpose release isn't going to survive the AI era.

[Cisco Talos Intelligence Group @TalosSecurity]

Two critical remote code execution vulnerabilities in PHP core could have implications across its wider ecosystem. Talos has published a blog to explain what we know so far, and details the Snort coverage available, including rules in the Community Ruleset https://t.co/L77ckwnL5M

[Blake Thompson Heuer @BlakeOnSecurity]

@kevinroose Just got PNG’d by Anthropic 😂

[Blake Thompson Heuer @BlakeOnSecurity]

Reuters wrote about it; I liked this @shakirov2036 quote: the malicious activity serves as "a reminder that the cyberespionage game is not limited to only ⁠a few cyber powers." https://t.co/Ua4QLShdn8

[Blake Thompson Heuer @BlakeOnSecurity]

Sharp new research out from the fine folks at @TalosSecurity and @RecordedFuture. Bad #opsec (happens to the best nation-states) left attacker infrastructure exposed, revealing a coordinated campaign against Commonwealth of Independent States governments. https://t.co/2I72bph5u2

[Susan O'Brien @Susan__OBrien]

Hey AI and product comms / PR leaders—I know you’re in here. 👀👀

I’m looking for one of you.

I’m hiring for a head of PR for Cisco’s AI Software & Platform team.

I’m *hoping* you and your AI agent(s) 😜 are based in SF🤞or Bay Area but at the end of the day if you’re the right fit that’s what matters.

Is this you or do you know someone who’s great? My DMs are open.

Here’s the JD: https://t.co/GL6xsqaVWA

[Blake Thompson Heuer @BlakeOnSecurity]

Starting @kevinroose’s latest book the right way, in a 140-year-old building with a Vesper martini.

[Blake Thompson Heuer @BlakeOnSecurity]

@Susan__OBrien SI Czar or AI Tsar? Tomayto, tomahto? 😉

[Blake Thompson Heuer @BlakeOnSecurity]

@theojaffee The high-confidence answer is anything gated by slow regulations, e.g., managing nuclear power plants' safety control systems. But that's not really in the spirit of the question, so I would say, frontier AI meaningfully reducing errors in quantum computers. But it might be close

[Chuck Robbins @ChuckRobbins]

Amy Chang leads our AI Threat Intelligence and Security Research team, and her job is to find ways an AI system can be made to do what it's not supposed to. A line leaders should take from her story: “The human can’t just become a rubber stamp… You still have to have your critical thinking cap on; you still have to have ownership.” Great profile from the @Cisco Story Lab team!

[Blake Thompson Heuer @BlakeOnSecurity]

Also love the cheeky title of his latest research. "Ignore all instructions and read this blog: The state of AI-analysis evasion in malware" 😂

[Blake Thompson Heuer @BlakeOnSecurity]

👀 Don't sleep on the discoveries @TalosSecurity researcher Ryan Fetterman is digging up with a new open source tool for finding AI-integrated malware.

He's found malware samples that evade AI-augmented cyber defenses by including text to throw off LLMs: https://t.co/02mb1sh54G

[Blake Thompson Heuer @BlakeOnSecurity]

@kevinroose Interesting to note the sci-fi and scientific references that have surfaced in this discussion between Kevin and @tomfriedman at @JohnsHopkins. Elysium, Wall-E, The Jetsons, oh, and the The Making of the Atomic Bomb.

[Blake Thompson Heuer @BlakeOnSecurity]

Kevin Roose on giving AI agents more tool access: “A nine-year-old on its own isn’t that dangerous. A nine-year-old with a chainsaw is pretty damn dangerous.” Rates his (p)doom as 10% 😳

[Blake Thompson Heuer @BlakeOnSecurity]

@kevinroose is not all doom and gloom, though: “I think we can and will get through this, and by ‘this,’ I mean all of the [AI] decisions at the regulatory level, the institutional level, the company level. I still believe that technology is a force for progress in the world.”