Zack Korman ✓

@ZackKorman · Oslo, Norway · joined 06 Jan 2014

Cofounder @ Embroidery. Building AI cybersecurity stuff.

16 605Followers

1 810Following

39 874Posts total

1.7MViews on collected posts

Against accounts of the same size

8 posts from the last 90 days, next to the 10K–100K follower range. shown far beyond its own audience, and the audience reacts.

Median views5 201this account1 266median for 10K–100K

Reach, %31.32%this account3.99%median for 10K–100K

Engagement, %2.66%this account1.98%median for 10K–100K

Metric	This account	Median for 10K–100K	Ratio
Median views per post	5 201	1 266	4.11×
Reach (views ÷ followers)	31.32%	3.99%	7.85×
Engagement rate	2.66%	1.98%	1.34×

Others in this range → Compare with another account → How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

56518 Aug

11.8K

4.6K19 Aug

588.5K

17.2K

9.3K

236.8K31 Aug

2.9K

2.1K1 Sep

6.8K

3.6K2 Sep

9.2K

1.1K

19.4K3 Sep

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

2.48%18 Aug

0.45%

0.26%19 Aug

0.82%

2.03%

1.69%

0.46%31 Aug

1.59%

2.83%1 Sep

3.04%

3.92%2 Sep

2.57%

2.75%

1.24%3 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.98%.

What the audience does

Likes70.1%7 902 in total

Reposts8.3%932 in total

Replies2.6%296 in total

Quotes1.2%138 in total

Bookmarks17.8%2 008 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

Followers by day

3 Sep

Daily snapshots since 03 Sep 2026; the dashed line is the starting count.

Latest posts

Totally normal post about AI safety / risk by one of the authors of the independent report on the OpenAI Hugging Face incident. https://t.co/vj0mUsOiRO 19.4K views · 168 likes · 20 reposts · 42 replies 03 Sep 2026 @ZackKorman We'd love referrals for great cybersecurity researchers/engineers! https://t.co/0IXhYotmmC 1.1K views · 27 likes · 0 reposts · 1 replies 02 Sep 2026 I'm sick of people saying "we need to improve cyber defense NOW", but their answer is always AI. That's not going to magically fix things. I have a full video coming tomorrow, but here's a short rant. https://t.co/WnGz5UjulD 9.2K views · 187 likes · 14 reposts · 31 replies 02 Sep 2026 I just want everyone to know that the point of my video was that I don’t support helping hospitals improve their security and I hope the misaligned cyber nukes escape containment. That was the message. https://t.co/T5jPXW3Oge 3.6K views · 116 likes · 9 reposts · 14 replies 02 Sep 2026 Oh nooooo the agents are spoofing tool calls again this transcript is so unreliable who knows maybe they escaped there’s no way to know noooo https://t.co/0YRSmkX1Am 6.8K views · 186 likes · 6 reposts · 14 replies 01 Sep 2026 @ZackKorman Respectfully, I'm pretty confused by your framing here, that this is some sort of turf war between AI safety / cybersecurity (eg 8:30 onward) I promise you, METR/Redwood are not trying to take cybersecurity's mantle or resources. You'll notice, for instance, that th 2.1K views · 57 likes · 0 reposts · 1 replies 01 Sep 2026 @ZackKorman Genuine question: Is your understanding that METR/Redwood were tasked with investigating the cybersecurity incident as a cybersecurity incident, or rather investigating OpenAI's model behavior *during the incident* as a question that is primarily one of alignment? e. 2.9K views · 44 likes · 0 reposts · 3 replies 31 Aug 2026 The independent review of the OpenAI Hugging Face incident, supposedly a watershed moment in cybersecurity, wasn't done by a cybersecurity firm and the authors have no cybersecurity experience. That's bad. Here's my new video. https://t.co/qYUyYp1apW 236.8K views · 821 likes · 140 reposts · 69 replies 31 Aug 2026 @ZackKorman this also allows you to bypass purview restrictions convos don't inherit sensitivity labels if you make copilot avoid citation 9.3K views · 148 likes · 4 reposts · 5 replies 19 Aug 2025 Here is a full write-up and why I think Microsoft is bad: https://t.co/luaZtN0Hwf 17.2K views · 296 likes · 37 reposts · 8 replies 19 Aug 2025 Microsoft isn’t disclosing this so: M365 Copilot allowed users to access files without producing an audit log. All you had to do was ask Copilot to not link to the file. You don’t even have to ask; it sometimes just happens. If your org uses Copilot your audit log is likely wrong 588.5K views · 4.2K likes · 562 reposts · 43 replies 19 Aug 2025 @ZackKorman I exposed a vulnerability in a video game once, they gave me the founders package. 😅 Never really thought about getting any further recognition other than the game being fixed. Your work made the world a safer place, albeit not by much (considering how many other v 4.6K views · 11 likes · 0 reposts · 1 replies 19 Aug 2025 @ZackKorman If it’s patched blog it and let the world decide? 11.8K views · 51 likes · 0 reposts · 2 replies 18 Aug 2025 @ZackKorman Sounds like the responsible thing to do is drop the PoC https://t.co/osOnRon8sW 565 views · 14 likes · 0 reposts · 0 replies 18 Aug 2025 Microsoft isn’t just not issuing a CVE, they’re actually not going to disclose this issue at all. 610.9K views · 917 likes · 65 reposts · 28 replies 18 Aug 2025 Microsoft now confirmed that because the vulnerability I reported is important, not critical, and because they’ve now fixed it they won’t issue a CVE. It’s like they actually want to discourage people from reporting. 140.9K views · 688 likes · 75 reposts · 34 replies 12 Aug 2025