A bit more at https://simonwillison.net/2026/Jan/12/superhuman-ai-exfiltra...
Superhuman AI Exfiltrates Emails
promptarmor.com
Superhuman AI Exfiltrates Emails
1–10 of 31 posts
Re: Superhuman AI Exfiltrates Emails
#2I'm impressed Superhuman seems to have handled this so well - lots of big names are fumbling with AI vuln disclosures. Grammarly is not necessarily who I would have bet on to get it right
Re: Superhuman AI Exfiltrates Emails
#3The primary exfiltration vector for LLMs is making network requests via images with sensitive data as parameters.
As Claude Code increasingly uses browser tools, we may need to move away from .env files to something encrypted, kind of like rails credentials, but without the secret key in the .env
Re: Superhuman AI Exfiltrates Emails
#4The primary exfiltration vector for LLMs is making network requests via images with sensitive data as parameters. As Claude Code increasingly uses browser tools, we may need to move away from .env files to something encrypted, kind of like rails credentials, but without the secret key in the .env
So you are going to take the untrusted tool that kept leaking your secrets, keep the secrets away from it but still use it to code the thing that uses the secrets? Are you actually reviewing the code it produces? In 99% of cases that's a "no" or a soft "sometimes".
Re: Superhuman AI Exfiltrates Emails
#5Programming used to prevent this by separating code from data. AI (currently) has no such safeguards.
Re: Superhuman AI Exfiltrates Emails
#6I'm impressed Superhuman seems to have handled this so well - lots of big names are fumbling with AI vuln disclosures. Grammarly is not necessarily who I would have bet on to get it right
I wonder how they handled it. Everybody's connecfing their AI to the Web, but it automatically means that any data AI has access to can be extracted by the attacker. The only safe way forward is to 1. disconnect the Web or 2. perhaps to filter the generated URLs aggressively.
Re: Superhuman AI Exfiltrates Emails
#7I'm impressed Superhuman seems to have handled this so well - lots of big names are fumbling with AI vuln disclosures. Grammarly is not necessarily who I would have bet on to get it right
I wonder how they handled it. Everybody's connecfing their AI to the Web, but it automatically means that any data AI has access to can be extracted by the attacker. The only safe way forward is to 1. disconnect the Web or 2. perhaps to filter the generated URLs aggressively.
We should have a clearer view of permissions of the AI, operations it does, and have one button per day to accept/deny operations from given data. Instead of auto approval.
Re: Superhuman AI Exfiltrates Emails
#8The primary exfiltration vector for LLMs is making network requests via images with sensitive data as parameters. As Claude Code increasingly uses browser tools, we may need to move away from .env files to something encrypted, kind of like rails credentials, but without the secret key in the .env
One tactic I've seen used in various situations is proxies outside the sandbox that augment requests with credentials / secrets etc.
Doesn't help in the case where the LLM is processing actually sensitive data, ofc.
Re: Superhuman AI Exfiltrates Emails
#9I'm impressed Superhuman seems to have handled this so well - lots of big names are fumbling with AI vuln disclosures. Grammarly is not necessarily who I would have bet on to get it right
I wonder how they handled it. Everybody's connecfing their AI to the Web, but it automatically means that any data AI has access to can be extracted by the attacker. The only safe way forward is to 1. disconnect the Web or 2. perhaps to filter the generated URLs aggressively.
Private data, untrusted data, communication: an LLM can safely have two of these, but never all three.
Browsing the web is both communication and untrusted data, so it must never have access to any trusted data if it has the ability to browse the web.
The problem is, so much of what people want from these things involves having all three.
Re: Superhuman AI Exfiltrates Emails
#10As limited as they are, LLMs are demonstrably smarter than a whole lot of people, and the number of people more clever than the best AI is going to dwindle, rapidly, especially in the domain of doing sneaky shit really fast on a computer.
There are countless examples of schemes in stories where codes and cryptography are used to exfiltrate information and evade detection, and these models are trained on every last piece of technical, practical text humanity has produced on the subject. All they have to do is contextualize what's likely being done to check and mash together two or three systems it thinks is likely to go under the radar.