Okay, so I get where the other side is coming from. The idea of a government having a key to your private messages feels invasive, and there's a real risk of abuse. Strong encryption protects journalists, activists, and regular people from hackers and overreach. That's a powerful point, and any system we design has to have serious, independent oversight to prevent misuse. But here's the thing: we already accept that in the physical world, with a proper warrant, law enforcement can search a home or tap a phone line. That's a cornerstone of justice. Today, end-to-end encryption creates a perfect, unbreakable seal. That means a warrant for a suspected terrorist's or a child predator's communications is just a piece of paper. It's useless. We're not talking about mass surveillance. We're talking about targeted access, under strict legal authority, for the most serious crimes. The alternative is creating spaces where criminal activity is completely invisible. How do we investigate a kidnapping plot coordinated on an encrypted app? How do we stop a terrorist attack being planned in a secret chat? We can't just throw our hands up and say "privacy wins, case closed." There's a real cost to that. The argument that a backdoor inherently weakens security for everyone is valid, but it's not an unsolvable technical challenge. It's a policy and engineering problem. We can design systems where access requires multiple, non-governmental keys or robust court oversight. The goal isn't to break encryption for everyone; it's to ensure that when there is a compelling, lawful need, we aren't completely locked out. So while I respect the absolute privacy stance, I think it ignores the practical reality of keeping people safe. A society needs both liberty and security. A carefully regulated, judicially approved backdoor mechanism is the imperfect but necessary balance. Without it, we're essentially giving criminals and bad actors a guaranteed safe haven, and that's a risk I don't think we should take.

But if the current tools are so effective, why are law enforcement agencies from democratic countries consistently raising the alarm? You mention metadata and endpoint monitoring, and that’s fair—they are useful. But when a suspect is using a fully encrypted platform, metadata just tells you who they talked to and when, not what they said. You can’t stop a bomb plot if all you know is that two people chatted at 3 a.m. I hear you on the cryptographer argument, and I take it seriously. Saying it’s just an “engineering problem” might sound glib. But we engineer solutions to impossible-seeming problems all the time. We secure nuclear codes. We protect financial systems. The idea that we can’t design a highly controlled, multi-party access system with extraordinary oversight seems like a surrender to fatalism, not a technical fact. Your strongest point is about authoritarian regimes. And you’re right—if we establish this principle, it could be abused by bad actors. But that’s an argument for building a system with democratic safeguards, not for doing nothing. We don’t abolish wiretaps because dictators misuse them; we create strict legal frameworks. Could a backdoor be leaked or hacked? Absolutely. That’s why the design and oversight would need to be the most robust system we’ve ever built, with transparency and independent judicial warrants. You say criminals will just switch platforms. But that’s a game of whack-a-mole we’re already losing. If all major platforms operating in a jurisdiction have the same legal obligation for lawful access, it shrinks those shadows. It raises the bar. So while I agree we shouldn’t burn down security for a shortcut, I think you’re underestimating the cost of the status quo. We’re asking law enforcement to fight 21st-century crime with 20th-century tools, and the gap is where real victims fall through. Isn’t the greater risk creating a world where the worst crimes can be planned in absolute secrecy?

Look, you're right to bring up WannaCry. That's a devastating real-world example of a powerful tool getting out and causing harm. And it's a sobering reminder that any system we build has to be designed with the absolute worst-case leak in mind. I'm not dismissing that risk at all. But I think we're talking past each other on a key point. You say we're betting on a technology that doesn't exist. I'd argue we're refusing to even try to build it because the theoretical risk feels too high. We engineer incredibly secure things for other vital functions—air traffic control, power grids, military communications. The demand for perfection here is being used as a reason for total inaction. Your point about criminals spinning up their own servers is fair, but it's also a bit of a red herring. Of course sophisticated actors can do that. But a huge amount of harmful activity—from local drug trafficking to child exploitation networks to urgent kidnapping threats—happens on mainstream, user-friendly apps precisely because that's where the victims and the convenience are. Making those platforms inhospitable to the highest-risk criminal planning does have a real impact. It forces operations into less convenient, more detectable channels. It's not about catching every single bad actor; it's about raising the cost and friction for the vast majority of harmful acts that happen in the digital spaces we all use. You ask who pays the price. I agree, it's ordinary people if the system is abused or fails. But who's paying the price right now, under the current system? It's the victim of a crime that couldn't be prevented because a plot was hidden behind an unbreakable seal. It's the family who gets a warrant for their missing daughter's messages, only to be told the technology makes it impossible. We're already making a trade. We've just chosen to accept the cost of total opacity. The core of my position isn't that this is easy or risk-free. It's that the alternative—a world where certain spaces are designed to be utterly lawless—is also a choice with profound costs. The goal isn't a magic bullet. It's creating a legal and technical framework where, in those rare, justified cases, we aren't completely powerless. Dismissing that possibility because we can't guarantee perfect safety means accepting a different, and in my view, more dangerous, kind of powerlessness.

Look, I have to push back on the idea that we're just accepting powerlessness. That's not what's happening. We have a system right now where law enforcement has a ton of tools, but there's this one locked door they can't open even with a judge's permission. You're framing this as if we've chosen lawlessness, but we haven't. We've chosen a specific type of privacy, and it's creating a real blind spot. Your strongest point remains the risk of leaks and abuse. WannaCry is the go-to example, and it's a good one. But that was a vulnerability stockpiled by a spy agency, not a designed access mechanism with built-in safeguards and oversight. We're talking about building something fundamentally different—a system where the "key" isn't held by one agency, but requires a court order and multiple independent parties to activate, with heavy auditing. The comparison isn't perfect. And on the idea that criminals will just move: sure, some will. But you're focusing on the tech-savvy kingpins. The reality is, a lot of crime is opportunistic and happens on the platforms that are easy to use. If a predator is grooming a child, they're going to use the app the child has on their phone. If a group is planning a local act of violence, they're using WhatsApp because everyone has it. Making those platforms not a guaranteed safe haven forces them onto less reliable, more trackable methods. It doesn't solve everything, but it raises the barrier and makes investigations possible where they currently hit a wall. You keep asking who pays the price. I agree it's ordinary people if we get this wrong. But we're already paying a price. We just don't see the crimes that weren't stopped, the kids who weren't found, the attacks that weren't prevented. That cost is invisible but no less real. The argument isn't that a backdoor is risk-free. It's that the current state of absolute, warrant-proof encryption carries a risk we've decided to ignore because the alternative feels scary. We regulate every other form of private communication in the physical world with a warrant. We don't say "well, safecrackers might learn how to pick locks, so we'll just let criminals have unbreakable safes." We develop better locks and better oversight. The principle is the same. Refusing to even try to engineer a solution because it's hard feels like surrendering a fundamental tool of justice. We can demand the strongest possible safeguards while still recognizing that total, unbreakable secrecy for bad actors isn't a tenable public safety outcome.

You're right that I haven't cited a specific case where breaking encryption was the sole barrier. But that's because in those cases, law enforcement hits the encryption wall and the investigation stops—there's no public report saying "we failed because of Signal." The absence of evidence isn't evidence of absence. And while metadata and traditional methods work sometimes, you're asking us to accept that they're always enough. That's a huge gamble. Your point about safes is interesting, but it actually proves my point. We don't demand a master key for every safe, but we do accept that with a lawful warrant, a locksmith or the police can drill it open. There's no technological guarantee of absolute secrecy. End-to-end encryption is different—it's a safe that can't be opened by any means, ever, even with a judge's order. That's a new and unprecedented category. On WannaCry, you're conflating a weaponized exploit stockpiled in secret with a designed, regulated access mechanism. I'm not asking for blind hope. I'm arguing for building a system with the key split among multiple independent entities, like judges, technical experts, and oversight bodies, where any access attempt is logged and audited. The NSA tool was a single point of failure created in darkness. What I'm proposing is the opposite—a system designed in the open, with multiple checks, precisely to avoid that scenario. You say the people who pay won't make the news. But the people paying the price right now are also invisible—they're the potential victims of crimes we can't prevent because we've decided total digital secrecy is untouchable. You're willing to accept that cost because the alternative feels risky. I'm saying we should channel our ingenuity into minimizing that risk, not surrendering to it. The fact that authoritarian regimes would abuse this isn't a reason for democracies to abandon the field; it's a reason to build a system with such strong democratic safeguards that it becomes a model of transparency and restraint, countering their opaque abuse. We already accept that no right is absolute. Free speech has limits. Privacy in your home can be breached with a warrant. Why have we decided that digital communications get an absolute, technological shield that overrides all other societal interests? That's the real recklessness—creating spaces where the rule of law simply cannot reach.