Three tracks
- Personal: AI that helps you run your own life: family, home, kitchen, money, health questions, travel, staying safe and keeping your messages in order.
- Business: AI that helps you run a business: selling on Shopify, Amazon and eBay, answering customer questions, and setting up bots and agents for a team.
- Professional: your own personal AI, used on public information only (companies, competitors, products, public profiles, leads, meeting prep), never your employer’s internal data, and always within your employer’s AI policy.
The one rule: public information only
Everything here uses information anyone can find: SEC EDGAR (the SEC’s free public filings database), government data and notices, company websites, press releases, published talks and reputable news.
Never paste into a personal AI:
- confidential or internal company data: documents, emails, chat messages, forecasts, pricing, strategy;
- customer data, CRM records or deal notes;
- your employer’s intellectual property: code, designs, product plans;
- passwords, account numbers or anything secret.
Internal-data tasks belong in your company’s approved tools.
First, check your employer’s AI policy
- Is there an AI policy? Read it; some employers limit personal AI use for any work task.
- Is there an approved AI tool? If so, use it first.
- Not sure? Ask your manager or the policy owner.
Check your assistant’s data settings too: personal ChatGPT and Claude both let you choose whether your chats help train models.
Researching people: where the line is
- Public professional information only: name, title, employer, announcements, published talks and articles.
- Nothing private: no home address, family, health, politics, religion or personal social media.
- No scraping: respect each site’s terms. LinkedIn’s User Agreement bans scraping tools, bots and copying content without consent.
- No pretexting or manipulation: never pose as someone else to get information or use what you learn to pressure anyone.
The cases
Each case has a public-sources workflow, a ready-to-copy prompt whose source steps the AIR-T checked in October 2026, and a Safety section.
- Research a company before a meeting or a new lead
- Build a competitor brief
- Find and qualify new leads from public sources, ethically
- Turn earnings releases, calls and filings into a meeting-ready briefing
- Compare products or vendors from their public pages
- Prepare for a meeting with a person, public profile only
- A weekly industry-watch digest
- Size a market from free government data
- Watch new federal rules and comment deadlines
- Scan public contract opportunities
- Map a technology area from public patents
Also in this track: team AI guardrails at work, deep research with cited reports, an earnings research agent and sales outreach research.
Always
- You stay in charge: check every fact and approve anything before it is shared or sent.
- Not professional advice: briefs are general information, not investment or legal advice.
Sources
- SEC, Search Filings: what EDGAR is.
- LinkedIn User Agreement: Dos and Don’ts.
- OpenAI Help, data controls: training setting.
- Anthropic, consumer terms update: training choice.
All accessed 2026-10-15.