In January, Department of Homeland Security (dhs) officers killed at least two people. In both cases, a federal agent grasped his gun, aimed it at a peaceful citizen, and shot them dead.
I learned that Google sells its Cloud services to the relevant agencies within dhs. I thought that was wrong. Federal agents should not be able to kill citizens in the street. I set out to find the most effective way to push my company to stop serving these agencies. My divestment campaign quickly broadened into an attempt to prevent Google from signing an unethical military AI deal, as the Pentagon started pressuring AI providers into military AI deals with no restrictions against use for killer robots or mass surveillance.1
I wanted AI ethics commitments to hold under pressure. In particular, I wanted Google DeepMind (gdm) to maintain its existing commitment against supporting killer robots. Over several months, I asked many people to act. I asked senior people—respected people—people with reputations silvered by their concern about AI ethics and safety. Nearly all declined.
Take Stuart Russell, a famous AI researcher who spent over a decade crusading against autonomous weapons. I worked at his lab for years. At a conference, on-stage, he agreed to push his organization to make a statement supporting AI providers against government coercion and promised a poll of its members. The statement and poll both vanished.
Or take Jeff Dean, who is Google’s Chief Scientist and the co-lead of Google’s Gemini AI project. In 2018, Jeff signed a pledge to never support the development or use of killer robots. I got Jeff to publicly and boldly co-sign an amicus brief (where outsiders weigh in to sway a lawsuit) backing Anthropic against the Pentagon. But I also asked him to use his immense leverage to stop Google from making its own unethical deal with the military, and I don’t think he did. He remains at Google despite his pledge.
I wrote a 25-page proposal containing contract language and oversight mechanisms. Military- and surveillance-law experts praised the proposal, which represented a principled counteroffer Google could have stood by. I sent the proposal to Demis Hassabis (gdm’s ceo) who routed it to senior policy staff, only for the proposal to wilt unattended until Google signed a deal.
Senior management had insisted that Google wouldn’t sign. I disagreed with them, but they largely ignored my warnings. While I may have increased the Pentagon’s hesitation around the deal, Google still signed a deal handing over their AI without restrictions against killer robots or mass AI spying. Google’s contract restrictions were even weaker than OpenAI’s. At that point, I couldn’t stay at Google in good conscience, so I left.
This essay tells the story of why I left Google DeepMind. It is also the story of something larger: how powerful people and institutions failed, one after another, to keep their AI ethics promises in the face of pressure.
On private communications
Throughout this essay, I never quote anyone’s private words without permission. Where private conversations matter, I characterize them minimally. I otherwise keep to my own actions, public information, and official communications. You can’t verify my characterizations, so weight them accordingly. This minimization has cut content which would have supported my arguments.
The stereotypical activist action is to make a petition. But Google had already ignored a large petition on this issue. Plus, Google’s executives likely hardened their company against stereotypical organizing tactics. Sit-ins, strikes, even a mass of Google engineers quitting: I deemed all of them ineffective (if I could even pull them off).
As I strategized, I judged that Google would not care about 100 random research engineers quitting. No, in the AI industry, talent is top-heavy and teams are driven by a few hard-to-replace stars. I didn’t need to coordinate 100 engineers. Perhaps I just needed to coordinate 10.
I’d followed the news and guessed that Sundar Pichai (Google’s ceo) was more of a businessman than a “make me a big speech about ethics and I’ll change my mind” kind of guy. But if a few hard-to-replace people were ready to walk, that would matter for the business, so Sundar might listen.
That’s when I remembered reading Jeff Dean tweeting about how bad ice was, retweeting Anne Frank quotes. Maybe I didn’t even need 10 engineers, I just needed one.
This is absolutely shameful. Agents of a federal agency unnecessarily escalating, and then executing a defenseless citizen whose offense appears to be using his cell phone camera. Every person regardless of political affiliation should be denouncing this.
January 24th, 2026
2458.4k
“Terrible things are happening outside. Poor helpless people are being dragged out of their homes. Families are torn apart. Men, women, and children are separated. Children come home from school to find that their parents have disappeared.”
Diary of Anne Frank January 13, 1943
January 25th, 2026
54834.8k
Jeff is considered a saint at Google. He was Google’s 30th employee, developed key algorithms, and is known as a man of principle. A common joke: it’s easier for Jeff Dean’s resume to list what he hasn’t achieved than what he has. He’s Google’s Chief Scientist and a co-lead of Google’s Gemini effort. A Jeff departure would be a disaster for the company.
But if Jeff cared so much and had so much leverage, why was Google in these ice contracts in the first place? Of course, you can’t be Chief Scientist and constantly get what you want by threatening to quit. But I still felt confused.
At first I thought about who could put me in touch with him. But (and this is a good general lesson) if you want to talk to someone about something, you can always just ask them!
I told Jeff that I respected him for speaking out, that I wanted Google to divest from the dhs supply chain. I asked if he shared these goals and, if so, how I could help.
He suggested it’d be reasonable for me to email a few guys. Their names: Sundar Pichai (ceo of Google), Demis Hassabis (ceo of Google DeepMind), and Thomas Kurian (ceo of Google Cloud). I thought “sure, Jeff. No problem. I’ll just tell them what I think.”
My email
I’m writing as a concerned employee at gdm. I recently messaged Jeff Dean regarding my concerns. He suggested that directly emailing the three of you was a reasonable next step.
I have no problem with Google working with lawful administrations of either US political party. My concern is not about politics, but rather about the events enabled by Google’s role in the dhs supply chain.
These are not standard, legitimate enforcement activities. These operations are troubling from a human rights perspective and also pose reputational risk to any vendors involved.
On 1/28/26, the dhs posted its 2025 AI Use Case Inventory, which lists Google as one of several GenAI providers which “improve the operational efficiency” of dhs. I urge Google to immediately stop working with ice (and dhs more broadly), including via support for third-party integrators facilitating these specific operations. Whether through a direct (“prime”) contract or through intermediaries (like ITC Federal), Cloud and Gemini must not power these operations.
History will judge the tech sector by its involvement in these events. I love working at Google, and I want to ensure Google is on the right side of that history.
Alexander Matt Turner
Research Scientist
They never replied. I returned to Jeff and asked for a lunch to discuss constructive opportunities for real change within Google. I told him: “any time, any place. I’ll drive down to Mountain View to meet with you.”
At this point, I thought this was where “plan A” would fail. To my surprise, he actually accepted, for a lunch a few weeks out.
The Pentagon threatened a private company with economic destruction. Usually, the government would say “no thanks, we will find another supplier who will provide terms we want.” In this case, the government threatened to falsely4 designate an American company as a “supply chain risk,” which would force all military contractors to stop using Anthropic.
I’d been following the Anthropic–Pentagon standoff for weeks. That morning, I read about the ultimatum. I was attending a conference in Paris held by the International Association for Safe and Ethical AI (iaseai), which is a nonprofit founded in 2024 to be “a unified voice” for safe and ethical AI. The world-famous AI scientist Stuart Russell chairs its steering committee. Its 2026 working groups include “Red Lines for Advanced AI,” focused on “autonomous weapons and escalation.” iaseai seemed built for a moment like this.
Iaseai’s venue would be full of influential AI professionals who care about ethics. I thought: we can organize a response, as a field, in support of 1) Anthropic’s right to do business without threat of destruction and 2) actual standards for whether and how to integrate AI into lethal autonomous weapons systems and surveillance apparatuses.
Anthropic had two days to comply with the administration’s demands. Perhaps other companies would agree to these “all lawful use” terms before the deadline. The main question which weighed on me: can I stop Google from caving, from accepting an “all lawful use” deal? If Anthropic says “no” and Google also says “no,” now we’re getting somewhere. That seemed hard. I wanted to make it happen anyway. Google could cave at any moment, whether before the Friday deadline or after.
The venue was the headquarters for the United Nations Educational, Scientific and Cultural Organization.
When I arrived at the iaseai venue, I expected some of the hundreds of AI professionals to be discussing the urgent ai ethics news which just dropped. Instead, no one besides me brought it up. People busied themselves with the usual abstractions: “how does public choice theory inform coordination problems?”.
Stuart Russell
Founder of iaseai. Co-authored the standard AI textbook used in over 1,500 universities. Founder of UC Berkeley’s Center for Human-Compatible AI, where I interned for several summers and completed a postdoc. For many years he was the only big-shot academic who took the existential risk from AI seriously. I had long appreciated that.
I briskly gathered information. Hinton was speaking remotely, so I’d have to reach him indirectly, likely through Bengio or Stuart. I knew Stuart already and had a friend who could get me in touch. Bengio was the wildcard. I saw him leaving the venue, so I ran to catch him.
“Yoshua, would you be willing to make a statement supporting Anthropic’s right to do business and pushing against unregulated killer robots?”. I clarified that his voice could influence the many gdm professionals who respect him, professionals who might be mobilized to push against “caving” to the Pentagon’s demands. He told me to email him.
Soon after, his office told me they decided they weren’t going to make a statement. They didn’t explain why not. Since I didn’t know Bengio, I figured I should focus my attention on Stuart Russell and then let him handle it.
At lunchtime, I saw Stuart speaking with someone 1-on-1. With the encouragement of several attendees, I overcame a dash of social anxiety and interrupted his conversation as politely as I could.
“Stuart, this is an extremely important situation and your voice matters,” I told him. “Can you make a statement? Can you get the iaseai organization to make a statement, too? Can you move the people inside of these companies?”. He considered. He agreed. He would try to get Bengio and Hinton on board. He would convene an iaseai vote asap and announce it at the closing of the conference later that day. He didn’t hedge his willingness to fight.
“Now that’s what I’m talking about,” I thought. “Here’s a powerful guy who knows his power.”
AI company / DoD deals mattered. They marked the first public, high-profile intersection between modern generative AI and military use restrictions. Whatever compromises (or capitulations) the companies made would reverberate as precedent into the future. Iaseai attendees had a chance to take their contributions from “abstract work which might inform policymakers” to “directly influencing precedent.” Stuart was now going to mobilize them.
As promised, he spoke about the issues at closing. I recorded his remarks during the Q&A, which were originally available on iaseai’s conference schedule but were later migrated and not reuploaded to iaseai’s official YouTube channel.
The question I asked Stuart (video of the full exchange)
The Department of Defense is threatening to take over Anthropic so they can use Claude without restrictions on lethal autonomous weapons or mass surveillance. What would you say to companies like Google and OpenAI who are still in negotiations?
Stuart Russell
I was going to mention this, and this is a topic on which we are likely to be polling all of the members of iaseai. To see if they would like iaseai to take a position on this issue.
[Stuart summarizes the situation.]
Should iaseai make a statement in support of Anthropic’s right not to have its software used in purposes that are outside the contracted areas that it agreed to have it used for?
The question said, “what about Google DeepMind, what about OpenAI”—are they also going to take the same stand as Anthropic? It seems like that’s up to them, but the right of a company to say “we don’t want to sell our product for that purpose”—and to be able to say that without threat of economic destruction. I’m not a lawyer, but it seems like that’s something that ought to be protected.
I was surprised. Stuart said “It seems like [taking the same stand] is up to them.” He didn’t encourage Google and OpenAI to avoid powering lethal autonomous weapons? Wasn’t that a defining cause for him over the last decade?
The next question from a conference attendee asked whether iaseai would take positions on current issues. As the session wrapped, iaseai’s interim executive director Mark Nitzberg took the stage.
Mark asks for a show-of-hands. “We will be asking the opinions of members… and if you’re not a member, go to the member desk.”A near-unanimous show of hands supporting an iaseai statement backing up Anthropic’s right to do business freely.
Many news outlets are discussing the extortion racket that the DoD is applying to Anthropic…
I think we will try to get an online poll if we have time to make this happen. So that we can publish a news release saying “92% (or whatever it turns out to be) of members of iaseai are in favour of the proposition that Anthropic should not be required to do what they don’t want to do.”
Stuart called the DoD’s actions an “extortion racket.” I appreciated the frankness, though frankness in the hall wasn’t the same as frankness in a statement the world would see.5 As I left, I signed up to be an iaseai member to participate in the poll. I thought it’d pass but wanted to make my voice heard. I paid the $75 membership fee.
Before the closing ceremony, Mark had texted me that the survey would close Thursday night. With 2/3 support, Stuart would sign a statement on iaseai’s behalf, and without it, Stuart would sign a statement personally with the option for others to co-sign.
Thursday morning greeted me. I didn’t see how to vote in the poll. Strange. I texted Mark. He said iaseai would have to act after Anthropic’s Friday deadline.
I told Mark that the situation was urgent: Google and OpenAI could move at any point. I urged him to talk to Stuart about making his own statement, any statement. I appreciated that Mark was helping me figure out this situation. Later that day, Anthropic indicated they would stick to the red lines in their existing contract with DoD.
(Thursday was supposed to be the start of a romantic vacation in Iceland. I didn’t really want to be doing any of this.)
Friday morning. Stuart had not made any public statement. I asked Mark, “What is possibly more important than spending a few minutes composing an email to his favorite reporter?”. From my conversation with Mark, I learned that neither Stuart nor iaseai would act.
But here’s the thing. Mark committed publicly at closing that iaseai would hold a member poll. Mark encouraged people to pay dues and become members to participate. Then iaseai leadership silently cancelled the poll.6 No statement ever came.
Mark gave me an evolving list of reasons for his organization not making a statement:
Iaseai needed more time to draft a statement and so would move after the Friday flashpoint.
Iaseai doesn’t know which principle it should cite for making a statement.
Iaseai worried that joining an Anthropic–Pentagon news cycle “has pros and cons.”
Mark was sympathetic but wanted to focus on refining iaseai processes for handling this kind of situation.
Iaseai didn’t need to make a statement anymore because Anthropic made a statement.
I could appreciate the first reason: after all, Stuart had repeatedly disclaimed “time permitting” for making a statement before the deadline. But that last reason in particular didn’t make sense to me. Anthropic made a statement saying they won’t budge, so iaseai doesn’t need to make a statement supporting Anthropic?
The Pentagon set the Friday deadline for Anthropic, not for iaseai. The Anthropic autonomous weapons issue had been in the news for weeks prior. Anthropic’s lawsuit against the Pentagon ran for weeks after. Google wouldn’t sign for two more months. Iaseai had plenty of time to act and plenty of ways to act, like filing an amicus brief, putting out a statement, or contacting senior decision-makers at Google.
OpenAI did announce a deal on Friday. OpenAI claimed its deal protected the same red lines against killer robots and mass surveillance that Anthropic had insisted upon. However, some analysts concluded that OpenAI’s contract language contains wide loopholes.
Over the next two months, I messaged Mark Nitzberg many times.7 I explained that I was trying to convince senior decision-makers inside of Google; that Stuart could reach out privately without the political cost of publicly opposing Trump; that even introductions would help.
A message I sent to Mark
March 30th, thirty days before the deal was reported as signed
While I wish iaseai’s decisions had been different, I would like to find a way for us to accomplish our shared goals.
Let’s rewind to Wednesday, when I had just learned of the Pentagon’s threat. I was operating under unknown time pressure. I had no idea whether Google was about to undermine Anthropic’s position by signing a deal. As I acted externally by lobbying Bengio and Stuart, I acted internally too.
[My message] Gemini’s “all lawful use” policy and the DoD pressure on Anthropic
I’m reading reporting that gdm is “close” to a deal to allow “all lawful use” of Gemini for classified purposes. Apparently, for unclassified work, Google already “removed some model-level restrictions” and agreed to “all lawful purposes” for DoD work.
Designate them a supply-chain risk, forcing all other military contractors to stop using Claude—“supply chain risk” is normally only used for compromised foreign firms!—and
Invoke the Defense Production Act to force Anthropic to provide Claude anyways.
I think it’s unacceptable to nationalize a lab to force them to provide lethal autonomous weapons and mass [AI profiling] tools. As an industry, we should stand strong with Anthropic. Our hard work should improve Gemini for our customers, not sharpen the abilities of lethal autonomous weapons systems.
I’m interested in others’ thoughts on this topic. How can we advocate effectively to ensure our AI Principles hold firm against this kind of external pressure for offensive military use without humans in the loop?
I began posting this kind of message in the gdm-only discussion channel. These messages received unusually strong and supportive engagement. I had several reasons for posting:
I wanted to raise the cost of silence from leadership. Remember, the whole point is stacking up enough cost to outweigh the benefits Sundar perceived from signing (like avoiding political retaliation from the Trump administration).
I often had strategic arguments which I wanted to tell my senior contacts. I was just a research scientist so it would be presumptuous for me to directly tell them “consider X.” However, I knew some of them read this channel, so I simply posted my arguments in the channel.
I wanted to create common knowledge among gdm employees that no, they are not alone in their discomfort with these contracts.
I pointed out the problems with the “protections” in OpenAI’s deal to inoculate gdm employees against similarly fake “protections” that Google might try to pass off. If employees didn’t buy the veneer, Google would pay a greater morale cost.
Later, some messages provided people a channel to hint they would leave Google if the deal passed, while still providing soft deniability.
The “” reacts also acted as a subtle organizing mechanism. My understanding was that Google frowned upon people directly organizing through large internal channels. I noted who responded with “” and reached out to them privately for the petition which I’m about to talk about.
I considered sending Jeff another DM: “please stop Google from signing the deal if you can.” But that seemed… not likely to do much. “What spurs people like Jeff to action?” I thought. Moral stakes: already present. Sense of ownership: not present. I couldn’t own the project because I had no direct power. As best I could tell, Jeff truly was one of the few people with both the power and inclination to oppose a deal. I wanted him to feel empowered and to know that gdm workers backed him up.
I wrote a petition to Jeff and, with the help of a few friends, got about 250 gdm / Research signatures in the next day or two. I will not reproduce the entire petition because I said it was “not public” and instructed people not to share externally.9 However, the New York Times did report on this petition, so some bits10 are already public:
“Please do everything in your power to stop any deal which crosses these basic red lines,” the employees wrote. “We love working at Google and want to be proud of our work.”
The Pentagon didn’t just designate Anthropic a supply chain risk. Pete Hegseth (the Secretary of the DoD) initially claimed that military contractors must stop all use of Claude, not just the use of Claude in their contracts. That impacted Anthropic’s major cloud and enterprise customers, so the designation looked deadly for their enterprise revenue. This requirement also seemed quite illegal, and Hegseth backed down on that part. (However, I think Hegseth’s tweet had already done damage by making businesses uncertain about whether they would be punished for using Anthropic.)
Anyway, on Monday or so, Anthropic challenged the Pentagon in court and asked for an injunction to stop the order. A nonprofit called Protect Democracy put together an amicus brief from AI professionals. The amicus communicated something like, “in our capacity as experts, we agree that Anthropic’s technical and policy concerns are legitimate, even though we work for competing labs.” I signed the document, got some colleagues11 to sign, and (with encouragement from an organizer) reached out to Jeff. He signed.
I was pleasantly surprised. Big move. (At this point, some of my friends started saying “based and Jeff-pilled” to describe things they approved of.)
But there is still reluctance among some Pentagon officials to rely on Google, two people familiar with discussions between the Defense Department and Google said. That’s because the company dropped a military contract in 2018 in response to protests from employees who argued that A.I. should not be used in weapons, the two people added.
Several top A.I. researchers at Google, as well as at OpenAI, also recently signed a legal briefing to support two lawsuits that Anthropic filed against the Defense Department. Anthropic is challenging the Pentagon for designating it a supply chain risk after it clashed with the department over how to use A.I. in warfare.
The participation of Google employees in the legal briefing added to some officials’ concerns about the company, reminding decision makers in the Pentagon and elsewhere of the past protests against the use of A.I. in weapons, one former official with knowledge of the discussions said. Some Trump administration officials are worried that even if Google agrees to have its A.I. used widely, the company could bow out again, the former official added.
In this time, I sent several memos to Google Legal suggesting that Google file an amicus too. I reminded them that if Google let the threat stand against Anthropic, the government would now have a gun to point at Google in all future negotiations: comply, or be labeled a supply-chain risk. In the end, Google didn’t file an amicus (but Microsoft did).
As I talked more with senior management, they kept arguing something like: “don’t worry. Leadership cares about these issues, too. They won’t sign an ‘all lawful use’ deal.” I considered the idea. Might it be true?
I stayed focused on my goal. I needed to move Sundar. Jeff could, I estimated, move Sundar. Did Jeff want to move Sundar? Would he walk from a company where he is revered, with a prestigious job seemingly tailored to him, leaving behind a quarter-lifetime of memories? On the other hand, every tech company in the world would want to hire Jeff, his legacy seemed secure, he already had who-knows-how-much money, and leaving would show principle.
I didn’t want to manipulate Jeff. I wanted to be real with him as one concerned person to another. I even avoided reading more than one article about him to keep myself from subconsciously tailoring my arguments to appeal to his psyche.
The way I saw it was: I don’t need to get Jeff to agree to quit over this. In a sense, he’d already agreed to quit over it. He had signed a pledge in 2018:
Signed by 5,218 people
The decision to take a human life should never be delegated to a machine. There is a moral component to this position, that we should not allow machines to make life-taking decisions for which others—or nobody—will be culpable. There is also a powerful pragmatic argument: lethal autonomous weapons, selecting and engaging targets without human intervention, would be dangerously destabilizing for every country and individual. […]
By removing the risk, attributability, and difficulty of taking human lives, lethal autonomous weapons could become powerful instruments of violence and oppression, especially when linked to surveillance and data systems. […]
We, the undersigned, call upon governments and government leaders to create a future with strong international norms, regulations and laws against lethal autonomous weapons. These currently being absent, we opt to hold ourselves to a high standard: we will neither participate in nor support the development, manufacture, trade, or use of lethal autonomous weapons.
Agreed. Mass surveillance violates the Fourth Amendment and has a chilling effect on freedom of expression. Surveillance systems are prone to misuse for political or discriminatory purposes.
As an American citizen, the last thing I want is government using AI for mass surveillance of Americans.
February 24th, 2026
February 25th, 2026
1684.3k
@JeffDean what’s your position on autonomous weapons?
February 25th, 2026
127
@TopherSpiro In 2018 I signed this letter and my position hasn’t changed.
Jeff Dean freely reiterated his pledge and agreed that “AI for mass surveillance of Americans” is “the last thing [he wants].”
I reasoned that if Jeff “will neither participate in… nor support the development… or use of lethal autonomous weapons,” then logically, Jeff would have to quit Google if it signed an “all lawful use” deal. I didn’t know if Jeff was actually ready to do anything like that. I wanted to prepare for the possibility.
Jeff, like all of us, is a human being. Humans tend to be worried by acting alone against a powerful entity. Even though I imagined Jeff’s weight would be enough to move Google on its own, I wanted him to feel supported. The first part of that was the petition I organized, signed by over 250 gdm employees asking Jeff to fight for them. The second part, though, was quieter.
I secured backup from several senior Google employees. If Jeff would agree to put his foot down with Sundar, they signaled that they would too. That said, the intended coalition was of size13 one: Jeff. I expected that Jeff’s influence would be enough on its own, as his departure could leave the Gemini project in shambles. Maybe Google would care more about its AI program than about retaliation from Trump.
I guessed that Sundar was skilled at defusing pressure from employees. I hoped this was a gambit he wouldn’t be able to stop, even if he knew about it.
Was Jeff a wrestler? I read an account of Demis trying to negotiate with Sundar for a separate legal structure for DeepMind. Any deal with Sundar seemed like it would need wrestling. Persistent wrestling. Furthermore, Sundar might try to defuse the tension with vague promises, so Jeff would need a specific proposal.
I mean, imagine Jeff and I sit down for lunch. Imagine he’s on board. Imagine he goes to Sundar and says “If we sign this deal, I won’t be able to stay at Google. Help me stay.” Imagine Sundar says “You’re important to me, Jeff. What would it take?”. What does Jeff say? “Don’t sign a bad deal” would probably lead to a bad contract that looks less bad on the surface.
I needed to draft contract language which would actually work.
Good red lines
Rule out the questionable use cases (autonomous targeting without human control, untargeted profiling) while allowing trustworthy ones like missile defense. Avoid the weaknesses flagged in legal analysis of Anthropic’s red lines.
Robust red lines
Cloud would push deals through any loophole, Google Legal seemed unlikely to tighten my drafting, and the Pentagon wouldn’t want terms at all. The language had to hold under pressure, with auditing that respected classification and operational security.
Minimal trust assumptions
I made the Chief Scientist the single root of trust that everything else hangs off of (I was betting everything on Jeff anyway). The Chief Scientist would staff a Review Body to advise on contracts.
Accountability via transparency
The Review Body would only privately advise Jeff and Sundar, but overriding it surfaces in a yearly transparency report to all AI employees. Dissolving it would require advance notice and disclosure of the exact outstanding non-compliance findings. I worked to ensure the Body couldn’t be defanged as quietly as Google’s 2018 principles were.
Minimal pain to opposed stakeholders
I gave Cloud 2 of 7 seats, recused staff only from their own deals, capped delays at 10 days, and protected deliberations under attorney-client privilege.
I considered a negotiations round. The counterparties would object to something, and I didn’t want that something to be load-bearing. Therefore, I included less-important provisions meant to be negotiated away, like “the Review Body can escalate to Alphabet’s Board via supermajority vote.”
In my personal time (taking vacation days), I created language that would be a good starting point. I composed 25 pages’ worth of “good starting point” material. I have published the Framework on my site in generic form applicable to any provider of both cloud and AI.
I sometimes call this “the Framework” as shorthand.
This Framework proposes two narrow Standards for the AI provided by The Company to government entities exercising coercive authority:
Human control over targeting and use of force. The Company’s AI won’t be used in systems that select and engage targets for force without appropriate human control over each engagement, evaluated on a use-case-by-use-case basis. Applies whether The Company provides the targeting system directly or simply provides AI components in a targeting pipeline. Includes a right to legal transparency regarding how systems will be lawfully deployed, with compliance verification conducted by a mutually agreed neutral auditor. Does not restrict anti-munition defensive systems, intelligence analysis subject to Standard 2, logistics, or R&D.
No untargeted AI profiling. The Company’s AI won’t convert bulk data into individualized intelligence on people who aren’t already specific, identified subjects of investigation. For all persons regardless of nationality, individualized AI-assisted analysis must be proportionate to the security interest served, may not be initiated based solely on demographic characteristics or political expression, and AI-generated outputs may not serve as the sole basis for initiating individualized scrutiny. Heightened protections for all persons in the U.S. regardless of status. Permits targeted analysis of identified subjects, aggregate research, and conflict zone analysis that improves noncombatant protection.
Transparency via yearly internal reports: An advisory seven-person Defense AI Review Body of senior staff, appointed by and reporting to the Chief Scientist.
Superseded by future laws: If Congress passes substantial legislation governing these usages, the Chief Scientist and Review Body (by supermajority vote) can retire one or both Standards.
The Company will not accept “all lawful use” as its standard. Where “all lawful use” language is demanded, The Company will require access to the legal memoranda establishing the lawfulness of intended uses. That transparency gives The Company the information to evaluate each use case against its Standards. Some will meet The Company’s standard. Some won’t. The ones that don’t, The Company declines.
With the help of TomSmith, I got feedback from experts in military and surveillance law. In particular, I got feedback from a foremost legal expert on human / AI warfighting integration. They said my Framework was “actually pretty good” and suggested improvements.
A friend drove me down to Mountain View to meet Jeff. Wearing a crisp dress shirt and slacks, I walked through the Gradient Canopy office. I had the Framework in my bag and I had a question on my mind: “What, if anything, does Jeff want to do?”. Was I going to meet a crusader? A bystander? A strategist?
A month prior, I asked for this lunch so we could discuss “constructive options” for making Google’s contract situation more ethical. Would he have plans of his own? Surely he could. He had so much more visibility and experience. I readied myself to toss my Framework aside and follow a better plan devised using more information. I didn’t care much. Even if Sundar adopted my Framework, I knew I was unlikely to be credited. That made me a bit sad, but I wanted to get the job done.
I had lunch with Jeff under the California sun. I proposed that Jeff head the potential Review Body, but he didn’t take up the idea. Beyond that, I won’t discuss the details of our lunch. I can only point to his public conduct: He tweeted and
” reacts.
and suggested improvements.
0 Comments
Log in to join the conversation.No comments yet. Be the first to share your thoughts.