Elephas masks names, emails and phone numbers on your Mac before a prompt reaches ChatGPT or Claude, on every plan including Free: Read more:

Elephas Smart Redaction masks names, emails, phone numbers and other identifiers before a prompt reaches ChatGPT or Claude, on every plan including Free. If you want to use AI safely, use it with Elephas.

The order does not mention privacy, data or user prompts. Your chats got no new protection. A day later, an FTC spokesperson confirmed an investigation of OpenAI, Anthropic and other AI companies under the FTC Act. The probe opened this summer, before the accord and the order.

The same day, Executive Order 14434 told federal agencies to say "Super Intelligence" instead of "AI" in official non-statutory documents. It defines the term as the technology the legal definition of AI already covers. Same technology, new label, for now.

Signers named by Forbes: Musk, Zuckerberg, Huang, Amodei, Pichai and OpenAI president Greg Brockman.

The four layers: internal controls, an internal team checking they work, an independent external auditor or evaluator, and an independent board committee. All four are about the model, not your data.

The White House Accord on Super Intelligence is a one-page voluntary pact signed Sept 29, 2026 by Trump and six AI company executives. It sets four layers of checks on AI models. It has no penalties, no deadlines and no duty to notify users. 🧵

AI is making it easier to create more. More drafts. More ideas. More prototypes. But creating more isn't the same as creating something worthwhile. As the cost of creation falls, judgment becomes more valuable.

Elephas redacts 28 types of sensitive data on your Mac before cloud models see them, and there is a free plan: Read more:

If you want to use AI safely, use it with Elephas. It works with ChatGPT, Claude and Gemini, and hides names and private details before anything is sent. Your private details stay on your Mac. Elephas starts at $19/month. Try Elephas for free.

A contract, a tax return or a medical letter goes up as a whole file. Names, account numbers and addresses go with it, unless something hides them first.

Projects hold 5 files on Free, 25 on Go and Plus, and 40 on Pro and Business. Custom GPTs hold 10 files. Storage is 25 GB per user and 100 GB per organization. The Free Library is 500 MB. The 80-file limit may be lower at peak times.

Your PDF can be well under 512 MB and still hit the 2 million token cap. A long file may upload fine and be only partly read. That is Elephas's finding, not an OpenAI statement.

ChatGPT file upload limit: 512 MB per document file and 2 million tokens per file. The token cap does not apply to spreadsheets. You can upload up to 80 files in a rolling 3 hours, and Free gets 3 uploads a day. 🧵

Whichever agent you pick, Elephas hides names and private details on your Mac before a prompt goes to ChatGPT, Claude or Gemini: Read more:

If you want to use AI safely, use it with Elephas. It works with ChatGPT, Claude and Gemini, and hides names and private details before anything is sent. 28 types of sensitive data are redacted on your Mac before any cloud call.

For confidential work, ABA Formal Opinion 512 (July 29, 2024) says lawyers should review a tool's privacy policy and terms to see how it collects and uses data, and get informed consent where needed.

Elephas's guide picks Perplexity for cited web research on a Mac with 24 GB or more, and Cowork for folder chores on non-confidential files.

Prices, per Elephas's comparison as of Oct 9: Claude Pro is $20 a month ($17 on annual billing) and Max starts at $100. Perplexity starts at $20 a month, and its Max is $200.

Perplexity launched Hybrid Compute for Mac on Sept 1, 2026. A local model handles private files and sensitive steps. It needs Apple silicon, macOS 15 or later and at least 24 GB of memory.