David Sacks on How Anthropic is Ironically Running Surveillance on their Latest Models
X (https://x.com/TheChiefNerd/status/2065810372817264941?s=46) 13 Jun 2026
Takeaways
- “Eight months ago, I said that Anthropic was engaged in a very sophisticated regulatory capture campaign.”
- “It’s not just the fact that they’re doing mandatory surveillance…They are now retaining for 30 days every prompt and every output you send to one of these Mythos class models. There are no exceptions. Even Enterprise customers who had signed zero data retention agreements—they do not have a choice.”
- “It’s not just the prompts and the output…it’s all the context you share with them. So, you know, all these agent platforms are basically storing all of your memories, all of your files, all of your data, and they’re passing them to the model in these giant context windows to get better responses.”
- “Anthropic is saying it will keep all of that, and it does it to build a profile on you, to classify you, and then to determine what capabilities it then unlocks.”
- “They would degrade the product…what they show you…if it decided, at Anthropic’s sole discretion, that you are not worthy of having access to that level of information.”
- “They would even do things like rewrite your prompt in the background.”
- “They would not tell you what they were doing. They would still charge you for the product that you thought you were getting, and they would never tell you that you were not getting frontier model capability.”
- “They are now saying they will disclose when they downgrade you, but they are still downgrading people when they decide that that person should not receive the appropriate information.
Hindsight / Insight / Foresight
- “I think what we’re getting here is a vision of where all this is headed.” (David Sacks)
- Putting aside the post’s veracity, one of the first principles I learned is, “To digitize is to surveil.” That principle has stuck with me and helped inform what I digitize—especially when it comes to what I enter into any text field. It’s on US as human beings to remain ever mindful and vigilant of the tech we use, especially and even more so when that tech is “free.”
Amazon CEO warned White House about Anthropic AI risks
LinkedIn (https://www.linkedin.com/news/story/amazon-ceo-warned-white-house-about-anthropic-ai-risks-7357780/) 13 Jun 2026
Takeaways
- Discussions between Amazon CEO Andy Jassy and U.S. officials led the Trump administration to ban foreign access to Anthropic’s AI models due to security risks.
- Jassy told officials, including Treasury Secretary Scott Bessent, that Amazon researchers were able to generate information “that could be used to aid cyberattacks” using Anthropic’s new Fable 5 model.
- In turn, the Claude maker abruptly blocked access to two of its powerful AI models, days after their public debut.
Hindsight / Insight / Foresight
- At least this helps explain why I haven’t been able to use it since yesterday.
- More importantly, it’s eye opening the way in which the head’s up was disclosed to the federal government.
OpenAI subpoenaed by state AGs over impact on users
LinkedIn (https://www.linkedin.com/news/story/openai-subpoenaed-by-state-ags-over-impact-on-users-8959698/) 13 Jun 2026
Takeaways
- A coalition of state attorneys general subpoenaed OpenAI on Friday, seeking information on a variety of company activities including advertising, use of consumer data and policies regarding minors and seniors.
- In a statement to The Journal, OpenAI said: “We take the concerns raised by state attorneys general seriously and intend to engage constructively with their offices.”
- OpenAI, which confidentially filed for an IPO last week, was recently sued by Florida’s attorney general over alleged safety risks from its ChatGPT platform.
- “AI is a new and powerful technology, and we work every day to safely bring its benefits to people in a responsible way,” the spokesperson said.
Hindsight / Insight / Foresight
- “This is an important signal of how quickly regulation is catching up with AI adoption. From a business and partnerships perspective, the bigger question is how companies turn compliance into a durable framework for trust, especially as requirements begin to diverge across jurisdictions” (comment from Angela W)
The next workers most vulnerable to AI are not in tech
LinkedIn (https://www.linkedin.com/news/story/the-next-workers-most-vulnerable-to-ai-are-not-in-tech-8969418/) 13 Jun 2026
Takeaways
- Economists are increasingly concerned that back-office workers, such as customer service representatives and payroll clerks, face significant risk to their jobs from AI adoption.
- These white-collar roles — often held by women — could become more vulnerable than tech jobs, which have been more prominent in the public discourse.
- The Times notes that they span industries, geographies, and companies large and small, and often pay high-school graduates a middle-class salary.
Hindsight / Insight / Foresight
• I wonder what AI solutions I can offer to help ensure the “AI displaced” earn a living. I’m not here to help everyone, but as part of my AI journey I am defining my own sense of revolutionary altruism. Does that altruism include educating the AI displaced on ways they can help themselves? After all, at this rate it could happen to anyone.