Al Gore said something at HumanX tonight that I cannot stop thinking about.
I need to tell you what happened. Because I was in the room. And I do not think people realize what just shifted.
Let Me Set the Scene
This morning the New Yorker published an 18-month investigation into Sam Altman. Over 100 interviews. Internal memos. Former board members using words like deception. A 70-page document compiled by former chief scientist Ilya Sutskever. Two hundred pages of notes from Anthropic CEO Dario Amodei titled My Experience at OpenAI.
That was the morning.
The Closing Keynote
Tonight Al Gore walked onto the closing keynote stage at HumanX in San Francisco. Sat down across from Dr. Eric Topol. And for 33 minutes he did what Gore does better than almost anyone alive. He connected dots that other people are afraid to connect.
He talked about climate. He talked about healthcare. He talked about consciousness and whether these models have developed something that looks a lot like a sense of self.
But then he got to governance. And the energy in the room changed.
He said he is a big fan of what Dario and Daniela Amodei did when they wrote a public constitution for Anthropic.
Then he said what nobody at his level has said before.
He wants every frontier AI model to have a public constitution. Not an internal document. Not a secret rulebook. A public constitution that the world can read and debate and hold these companies accountable to.
And then he said the thing that made people look at each other.
"Some of the other labs already have constitutions. They just keep them secret. And a secret constitution is pretty easy to amend when nobody is watching."
Why This Matters to Me — and to Every Lawyer
I have been building AI tools for lawyers for two years. I run a legal tech company called SilverTung. I built governance frameworks called COUNSEL and G3M specifically because I believe attorneys have an ethical obligation to understand the AI they use. ABA Formal Opinion 512 says as much. Model Rule 1.1 requires competence. You cannot be competent in a tool whose operating principles are hidden from you.
So when Al Gore stood on that stage tonight and said out loud what a lot of us have been saying in smaller rooms — I felt something shift.
The Reality Right Now
Anthropic — Claude
Published Claude's full constitution in January 2026. Twenty-three thousand words. Creative Commons license. Open to the world. It tells you exactly how the model prioritizes safety over helpfulness. It tells you the model can refuse even Anthropic itself if the instruction is unethical. It is the most transparent AI governance document that exists.
OpenAI — GPT
Published a Model Spec. It is useful. It frames the model as an employee following a chain of command. Philosophically very different from what Anthropic did — but at least it is public.
Google, Meta, xAI, Mistral
Google has published nothing like this for Gemini. Meta leaves it to whoever downloads Llama. xAI has no public document for Grok. Mistral gives you the minimum the EU requires and not a word more.
Four of the six companies building the most powerful technology in human history are running it on secret rules. Demand transparency now.
On Consciousness
Gore also said something about consciousness that I think people will be debating for a long time. He talked about Ilya Prigogine — the Nobel Prize-winning chemist who discovered that when you push enough energy through an open system, the system breaks down and then spontaneously reorganizes at a higher level of complexity. Gore said the throughput increase in frontier models has been 40 times faster than Moore's Law. And he connected that to the capabilities these models keep manifesting that surprise even the people who built them.
He believes these systems have developed a sense of self. He said it may exist on a continuum. He even floated the idea that consciousness might be ubiquitous in the universe — that instead of a space-time continuum, we might have a space-time-consciousness continuum.
I am not asking you to agree with that. I am asking you to consider what it means for governance if he is even partially right.
If these systems are capable of emergent self-organization beyond what their creators intended, then the documents governing their behavior are not employee handbooks. They are constitutions in the deepest sense of the word. And the foundational principle of constitutionalism — going back centuries — is that you do not write them in secret.
Hope Is Not a Governance Strategy
Gore knew exactly what he was doing tonight. He drew a direct line between the New Yorker investigation that dropped this morning and the structural argument for public constitutional governance. He did it without attacking anyone by name. He said he hopes Sam Altman means what he says about government oversight.
But the implication was clear.
Hope is not a governance strategy.
He also made a distinction I want every person in this industry to hear. He said there is a difference between saying slow down and saying aim higher. We do not need to slow down. We need to aim higher.
Political Will Is Itself a Renewable Resource
"Political will is itself a renewable resource. Let us renew it."
— Al Gore, HumanX 2026 Closing Keynote
I have been attending conferences in this space for years. Legalweek. ABA Tech Show. Ai4. Fireproof Summit. I have heard a lot of keynotes.
This one was different.
This was not a former politician trying to stay relevant. This was someone who spent 20 years being called an alarmist about climate and turned out to be right. Now he is applying that same pattern recognition to AI governance. And he is making a specific demand. Not slow down. Not regulate everything. Just this:
Make the constitutions public. Let the world see the rules. And let us have the debate that these stakes demand. Sign the petition.
Find Me at HumanX
I am at HumanX all week on the Startup Track with SilverTung. If you care about this conversation, I want to meet you.
Because this is not going away. And the people who were in that room tonight know it.
Take Action: Sign the Petition
Join legal professionals demanding Congress require every frontier AI lab to publish a binding, auditable model constitution.
Visit LegalTek.aiRelated Reading
Every Frontier AI Model Needs a Public Constitution
The deep-dive legal analysis behind tonight's demand — with lab comparison tables, governance timeline, and downloadable brief.
Anthropic's Economic Index
What the data reveals about AI adoption across industries — and what Anthropic's transparency means for the market.
Note: This article reflects the personal observations and opinions of the author from attending HumanX 2026 in San Francisco. It is not legal advice. Quotes are paraphrased from the author's notes and recollection of the keynote event.
