The Hackening

Open questions

Things nobody has a confident answer to yet. Each question shows what is known so far and where it comes from. Where an answer gives a view or an estimate, it says whose it is.

Could AI-driven hacking shut down gas, electricity, water, roads, trucking or the police?

Current best answer
Nobody knows. Water and sewer systems look unusually exposed, because they are split among thousands of small operators that must each find and fix their own flaws. On Manifold, a prediction site that uses play money, a market with about 40 participants put about 6% on a major US blackout caused by a cyberattack before 1 January 2027 (as of 14 Nov 2026). A market that small is a rough signal, not a forecast. One commenter, Thane Ruthenis, argues that what matters more is how fast emergency services could restore a basic offline service: power and deliveries worked in the 1970s, and Ukraine has kept essentials running under bombardment. Another, Archimedes, suggests expecting regular disruptions at least as bad as the November 2026 Outlook outage and US flight halt, without saying those were attacks.

Does anyone able to cause long outages actually want them?

Current best answer
Nobody knows. One commenter, Jenna_Parr, sorts attackers into three groups: criminals, who want money and so favor theft, extortion and ransoms; states and people who want destruction for its own sake; and AI agents that no longer answer to anyone. On that view, lasting damage is more likely to come from the last two. But criminals have caused long outages too: ransomware kept one US credit union's online banking down for about two weeks in 2024.

What would an attack on many banks at once actually look like? Where would the money go, and couldn't the government block large transfers?

Current best answer
Nobody has a clear picture. Most reported attacks on banks so far have leaked data, but money has been moved too: about €100 million was stolen from the French bank Crédit Mutuel Arkéa in November 2026, out of the bank's own account. Another way many banks could fail together is customers losing access when a supplier that many banks share goes down, as in the 2023 Trellance attack. Whether a government could block very large transfers has not been answered.

Do we get a "death spiral" where infrastructure goes down faster than it recovers, or do we muddle through with scattered outages and theft?

Current best answer
Nobody knows. Thane Ruthenis, who asked, was unsure: it "seems possible" there will be a spiral, "or maybe not". This site's maintainer guesses the chance of a months-long, hard-to-reverse breakdown of online finance at about 1–3% over 18 months (a guess made in late October 2026). That is one person's judgment, not a forecast. The reasoning: past attacks on financial firms caused outages of days to weeks followed by recovery, and banks keep offline backups of account records. Against that, nothing like today's tools has been seen before, so the past is a weak guide. On 14 November 2026 Thane Ruthenis added that the real question is how quickly basic services could be restored offline, and that preparing for total breakdown "doesn't really seem threat-appropriate"; he also asked whether he was underrating how fragile these systems are.

Sheltered Harbor, which sets the standard for banks' nightly offline backups, is closing. Will banks keep the backups, and will anyone independent still check them?

Current best answer
Not known. The maintainer's guess is that the backups continue, because banks and their suppliers run the storage themselves. The closing notice hands the standard to the industry group FS-ISAC and doesn't mention independent certification.

Are some password managers safer than others, and is a good one good enough?

Current best answer
Nobody in the discussion has answered yet. Elizabeth asks whether one well-known password manager is better than another, naming LastPass and 1Password, and whether the better one is good enough. This site has not compared them or checked either one's record. Its guide says the free password managers built into Apple and Google devices are enough to start with. Elizabeth also points out that replacing old backup codes (the one-time codes that let you into an account without your phone) seems worth doing but was on nobody's list; the backup guide now includes it.

How many days of water, food and supplies are enough, and how much extra should you keep for neighbors who didn't prepare?

Current best answer
Most guides plan for three to seven days; the Red Cross figure for water and food when staying at home is about two weeks. In the best-known hack of a single financial institution, a California credit union in 2024, online banking was out for about two weeks. One commenter, Philipp Risius, suggests planning for up to two weeks without mains electricity. Another, Thane Ruthenis, doubts that preparing for a total breakdown fits the threat. Nobody has answered how much to keep for neighbors.

When would you actually use a survival AI chatbot that runs on your own computer without the internet?

Current best answer
Probably rarely, in the view of most of the commenters who have discussed it; none says they have tried one in an emergency. The useful part is the offline reference library it searches, which is free on its own. One commenter, lilkim2025, expects a plain search tool over the same manuals to be faster and more reliable than a small chatbot. Another, Thane Ruthenis, thinks it is not unhelpful but doesn't fit the threat well. Most of what you'd need is specific to your situation and has to be prepared in advance.

If attackers and defenders have equally capable AI, don't defenders win, since they know their own systems and can outspend any one attacker?

Current best answer
Not automatically, and people who follow this disagree. Finding a flaw first isn't enough. Defenders found the PHP2Shell flaw on August 1, yet attackers were using it by October 17, four days before a fix was released. The writer Zvi called that "an update against defenders being advantaged". Grendel1209, who described working in cybersecurity at a threat-intelligence company, expects freely downloadable AI models to favor attackers. Vitalik Buterin predicts defense wins in the end, "but the transition will be hard". CrowdStrike, a security firm, reported that in the first half of 2026, for flaws with public demonstration code, 88% of the attacks it observed came within 48 hours of that code's release.

When will AI models that anyone can download be as good at hacking as the most advanced ones?

Current best answer
Two reports in September 2026 give part of the answer. CAISI, the US government's AI testing center, said on 17 September that GLM-5.3, a Chinese model anyone can download, was the most capable downloadable model at hacking so far. It also said GLM-5.3 was "significantly lower" than the best US models, about four months behind on its tests. Anthropic, an AI company, reported on 29 September that in its tests GLM-5.3 built working attacks about as often as its own model Claude Mythos Preview, announced about five months earlier: 50 of 410 attempts against 56 on one test. Anthropic also found GLM-5.3's safety limits easy to get around. So on these tests downloadable models have reached about where the leading models were a few months before, and they still trail the newest US models, some of which are released only to vetted users. Tests are not real-world attack ability. Earlier guesses in the discussion were slower. In August 2026 hadad put 85% on a downloadable model hacking as well as Mythos, a leading model, within 24 months, and Vladimir_Nesov guessed "maybe 2 years, in 2028". In October 2026 the writer Zvi judged the best downloadable model "more than six months behind". Forecasters at the forecasting site Metaculus put about 95% on one matching Mythos on a hacking test before July 2027.

Against AI-driven hacking, is your money safer at one of the five biggest banks or at a credit union?

Current best answer
Disputed, and nothing here is a recommendation to move your money. Grendel1209, who described working in cybersecurity at a threat-intelligence company, recommends a top-five bank, on the view that only the biggest firms can afford strong defenses. Thane Ruthenis suggests splitting money across several different banks. The maintainer's own reasoning, which is untested, is that a big bank and a credit union may not fail at the same time, so holding accounts at both is the cautious choice.