Last updated: August 2026
Trust, transparency, and responsibility are core values that guide my work, whether through consulting, writing, speaking, or creating content for my newsletters, social media and podcast. I believe readers deserve clear insights into how my content is produced, the sources of my income, and the ethical standards I uphold. This statement outlines these practices in detail to ensure transparency and accountability.
How I Earn Income
My professional work includes independent consulting mainly for technology companies and non-profit organizations. This consulting primarily focuses on democracy and technology, trust and safety, and the intersection of technology with society and politics. I also offer training sessions, facilitate workshops, and participate in speaking engagements at various conferences and events. I cannot disclose every client due to NDAs, but where I can, I will, and I will always require disclosure and labeling of any sponsored content I get versus my own analysis. I also earn income from paid subscriptions, the Anchor Change Briefing Network, and the Panic Responsibly merch store.
Sponsorship Policy
Whenever I receive compensation from an individual, company, or organization for content featured in my newsletter, this will be identified as sponsored. Additionally, I retain full editorial control over which sponsored material I accept to ensure it aligns with my values, accuracy standards, and my readers’ interests.
Use of Artificial Intelligence Tools
I frequently use AI tools in my creative, research, analytical, and editorial processes, primarily ChatGPT, Claude, Gemini, Grammarly, and increasingly other tools as their capabilities evolve.
The most important thing to understand about how I use AI is that I remain responsible for what I publish. AI can help me think, surface connections, organize, draft, edit, analyze, and package my work. But I independently interrogate and verify what I use, make the final editorial decisions, and stand behind everything that goes out under my name. “The AI did it” is not an excuse for getting something wrong.
Ideas, Brainstorming, and Analysis: My work starts with what I observe, the questions I have, the experiences I’ve had, and the ideas and connections I think are important to share. I use AI as a thought partner to brainstorm, ask questions, challenge my thinking, identify gaps, and sometimes surface ideas or connections I hadn’t considered. When AI introduces something new that I want to use, I independently verify it and ensure I appropriately credit other people’s work.
Voice, Brain Dumps, and Drafting: Increasingly, I talk through my ideas rather than typing them all out. Sometimes that simply means using a tool like Wispr Flow to dictate rather than type on a keyboard. Other times, I use voice conversations with tools such as ChatGPT or Claude to talk through an experience or idea and have the AI ask me questions that help draw out my thinking.
I may then add additional thoughts, notes, research, or instructions and ask AI to turn that material into a first draft. I edit and revise that draft myself and may put it back through one or more AI tools for additional feedback and refinement. The amount of AI involvement varies by piece.
Research, News, and Summaries: I use AI to help me orient myself to topics, synthesize information, summarize articles and other source material, and increasingly to monitor and summarize news. Sometimes I use AI with web search enabled, and sometimes I work only with material I’ve provided.
For article summaries and other material that I publish or rely upon, I go back to the underlying sources. I read the articles I include, check summaries against them, and decide for myself which information matters. Where appropriate, I link to the underlying material so readers can examine it themselves.
Polling and Data Analysis: I use AI as an additional analytical tool when working with polling and other data. My typical process is to review the toplines and crosstabs myself first and identify what I find noteworthy. I may then give the data to AI and ask it to look for patterns, connections, or findings I may have missed. If it surfaces something interesting, I independently interrogate and verify it before using it. I double-check numbers rather than assuming an AI-generated analysis is correct.
Editing and Refinement: I use AI as an editing partner to identify unclear sections, refine wording, suggest cuts or alternative approaches, and point out things I may have missed. I may compare suggestions from multiple models. I selectively use those suggestions rather than automatically accepting them. I also use Grammarly for grammar, punctuation, readability, and copy editing.
Titles, Packaging, and Social Media: I use AI to help generate and refine titles, subtitles, SEO titles and descriptions, URLs, newsletter packaging, and social-media content derived from my work. I may also use AI to help format newsletters into my standard templates.
Transcription and Recordings: I use AI transcription for several purposes. I dictate prompts and ideas when speaking is more convenient than typing. I may transcribe meetings with permission to accurately capture what was discussed and use those notes in follow-up work, including proposals. I also sometimes record and transcribe public talks, conference sessions, webinars, and similar events so I can refer back to that material later.
Images and Graphics: I occasionally use generative AI to create illustrative or conceptual images, and I identify them as AI-generated. I also use tools such as Canva’s AI features to help create or refine graphics and other visual assets for social media and my content.
Coding and Client Work: I am experimenting more with AI-assisted coding and “vibe coding,” including for some client work. When I use these approaches for clients, I am transparent with them about what I am doing and the tools I am using.
Agents and New AI Workflows: I am experimenting with AI agents and other emerging workflows, particularly for tasks such as monitoring and summarizing news. I don’t think every minor use of AI needs to be cataloged here. I will update this statement when a new tool or workflow materially changes how my work is researched, created, analyzed, or published.
Information Shared with AI Models
My personal writing, notes, ideas, and other content regularly go into AI systems because I find them useful for my work. Where providers give me controls over whether my content is used to improve or train their broader models, I have turned those settings off.
When working with clients, I do not enter confidential or proprietary client information into general AI tools unless the information is already public or the client has authorized its use. I work with clients to understand their AI policies and preferences, including whether they provide secure internal AI tools for collaborative work.
I also avoid intentionally sharing sensitive personal information such as my home address, phone number, or other private details when it isn’t necessary.
Accountability, Verification, and Attribution
I am responsible for everything I publish.
I double-check factual claims, numbers, and quotes rather than assuming AI has gotten them right. When I cite or rely on other people’s reporting, research, ideas, or work, I try to link to the original source so I can verify it and readers can examine it themselves.
AI sometimes gives me a connection, phrasing, or idea that I hadn’t considered. I am comfortable pursuing those contributions when I have independently interrogated them, verified what needs verifying, and satisfied myself that I am not presenting someone else’s work as my own.
The ideas I choose to pursue, the observations I think matter, the conclusions I draw, and the final decisions about what gets published remain my responsibility.
How I Disclose AI Use
This statement describes my normal use of AI across my writing, research, analysis, consulting, and content creation. I don’t add a separate AI disclaimer to every newsletter or social-media post because doing so would quickly become repetitive without giving readers much additional information.
When AI plays a materially different role in a particular piece than the practices described here, I will disclose that separately when I believe that difference would be meaningful for readers to understand. AI-generated images will continue to be identified as AI-generated.
My practices will continue to evolve as these technologies change. I’ll update this statement when my use changes enough that this description no longer accurately reflects how my work gets made.
Privacy and Data Handling
I respect my subscribers’ privacy. I collect minimal personal data, primarily email addresses, solely to deliver newsletters, manage subscriptions, and occasionally conduct reader surveys to improve content relevance. I do ask for your address with the intake form sent to you when you sign up. I promise to use it only sparingly to send fun surprises. If you ever want me to delete your address, just shoot me a note. Subscriber data is never shared, sold, or rented to third parties. Any analytical tracking is anonymized and used exclusively to enhance reader experience and content quality. Stripe is the payment processor for paid subscription options.
Affiliate Links and Commissions
Occasionally, I may include affiliate links to products or services I genuinely recommend or have personally vetted. If readers purchase through these links, I may receive a commission at no extra cost to the reader. Recommendations are based solely on merit, personal experience, or careful assessment unless otherwise labeled as a sponsorship.
Corrections and Accountability
Despite diligent processes, mistakes happen. Corrections are promptly and transparently communicated to readers when errors are identified, whether factual inaccuracies or omissions. I welcome reader feedback and hold myself accountable to the highest ethical standards.