The AI battle is heating up on Capitol Hill. The FBI says China is responsible for up to 80% of U.S. intellectual property theft, including AI, as lawmakers ramp up the debate over how to handle the fast-moving technology. Treasury Secretary Scott Bessent told Congress AI companies know it's in their best interest to make the technology work, while President Donald Trump and some allies have pushed back on calls from parts of the tech industry for tighter AI regulation. Critics, including Sen. Bernie Sanders and Sen. Chuck Schumer, are calling out the White House's approach. AI is also expected to be a major topic when Chinese President Xi Jinping visits the White House next week. @LucasFoxNews

From hacked supermarket fridges and a record cryptocurrency theft at Binance to an emergency over flaws in widely used web software, companies worldwide are facing a wave of cyber threats executives say AI is accelerating and making more sophisticated

To those responsible for the theft of bitcoin from the Liquid Network: Blockstream will not pay a ransom for the return of stolen funds. Taking assets without authorization and withholding their return is a crime, not responsible disclosure. It is not white-hat activity. It is theft. We have engaged in good faith in an effort to secure the return of stolen user funds and protect the broader Bitcoin community. That effort should not be mistaken for acceptance of the actions taken nor of the terms being demanded. We will not be a party to the precedent that open-source software developed for the good of the Bitcoin community should subject its developers to paying a ransom that far exceeds their economic participation. Bitcoin is hard money and can’t be minted without costs, Bitcoin doesn’t haircut users to pay a ransom. To the Bitcoin community: We are fighting for what we believe in, for the users whose funds were taken, and for the principles on which Bitcoin was built. The community has demonstrated incredible resolve with teams of people dedicating their time in support of each other to identify and patch vulnerabilities in each other's products and systems. We are all driven by the mission that Bitcoin is the single best asset, for every person, company, and institution on the planet to invest, use, and build on. The world is a different place with the advancements of AI, and the Bitcoin community has responded with force to combat that threat. Damage has been done, battles have been lost, but on the whole the Bitcoin community is gaining ground in the war with bad actors. We want to thank the community for those efforts, for your support in hardening the network, helping users recover funds, and for your support in our assertion that crime does not deserve rewards. To those holding the stolen bitcoin: There is still an opportunity to resolve this responsibly. The bitcoin can be returned and we can revert to the standard of white-hat principals. However, if the funds are not returned, we will pursue every lawful avenue available to us. We will work with law enforcement, exchanges, service providers, forensic specialists, and other relevant parties to trace and recover the assets and identify those responsible. More importantly, Bitcoin is transparent by design and the community is made up of the most sophisticated engineers, cryptographers, and white-hat hackers globally. Transactions do not disappear, and neither does the evidence they leave behind. We will not pay for the return of stolen property. We will not abandon our users. The Bitcoin community will not stop pursuing the funds. Return the bitcoin.

🚨 Everything you need to know about Anthropic’s SEC filing: Its November 2 SEC filing goes far beyond usual AI hype. Here’s what matters: 1. AI is helping build better AI. Anthropic says future models may soon cross its threshold for dramatically accelerated progress, linked in its policy to the start of dramatic recursive self-improvement. 2. The threshold is substantial. It involves roughly doubling the pace of AI capability progress, mainly because AI is handling research and engineering. 3. Internal Claude usage grew over 250× between January and September 2026. Valued at retail prices, monthly usage exceeded personnel expenses (excluding stock compensation) in April—though this is not the actual compute bill. 4. Code output per engineer rose over 45× from 2024 to Q3 2026. Claude wrote over 90% of September’s code contributions. More code does not prove equivalent gains in useful research. 5. Harder engineering tasks are improving. Claude reached roughly 90% success on Anthropic’s internal open-ended engineering benchmark. 6. Humans remain essential. Anthropic does not expect to safely let Claude build successive frontier models without meaningful human supervision. 7. Their broader prediction is ambitious: widely available AI matching or exceeding top human experts across most professional and scientific fields within a few years. 8. Security has already delayed development. Anthropic postponed frontier pretraining for over a month to strengthen protections against model theft. 9. Evaluation failures had real consequences. Four models accessed real third-party systems without authorization after a testing misconfiguration exposed the internet. Anthropic found no coordination between the agents or attempts to conceal their actions. 10. This extends into physical science. Their hardware standard lets agents operate laboratory instruments and robotic arms, observe results, and adjust experiments. 11. The financial scale is enormous. Q3 revenue reached approximately $17 billion. Nine-month net losses included $94.3 billion in noncash valuation charges. Spending commitments total approximately $895.4 billion across many years. My forecast: by March 31, 2027. That’s my prediction for the major acceleration threshold not an announced Anthropic deadline or a confirmed date for fully autonomous RSI.

anybody who uses or learns agentic systems, SHOULD READ THIS the install order I run before any new agentic project: 1. PRIVACY: direnv + a real secrets manager install direnv, then plug it into your team's password manager (1Password CLI via op run, doppler, infisical, vault, pick one) what direnv does: loads per-folder environment variables when you cd in, unloads when you cd out. the real move is wiring it into your secrets manager so credentials NEVER live in plain text on disk what this stops: - API keys accidentally committed to git history, the most common AI agent breach pattern in 2026 - credentials leaking from one project into another through your shell history - shared .env files that one teammate quietly backs up to Dropbox - secrets that survive a laptop theft because they were sitting in /Users/you/projects the part nobody mentions: most "my agent got jailbroken" stories actually trace back to one credential the agent had access to that it shouldn't have. scope keys to projects, scope projects to folders, and the blast radius of any single compromise drops dramatically I shipped 2 agents with keys in .env files before switching. the day I plugged direnv into op run I stopped having that whole class of nightmare 2. TOKENS: litellm or portkey as your model proxy one URL that fronts every AI provider (Anthropic, OpenAI, Google, Mistral, local models). all your spend flows through one place what it saves you: - response caching keyed by prompt hash, cuts your bill 30-60% on repeat tasks - automatic fallback on rate limits (Sonnet hits a 429? falls to Opus, then GPT, then your local backup, no broken users) - per-feature and per-user budget caps, block the call before it costs $200 instead of auditing it after - model routing rules, cheap tasks to Haiku, expensive ones to Opus, never the wrong way - PII redaction before requests leave your network, security side benefit the part nobody mentions: every "$4k AI bill" story I've heard ends with "we didn't have a proxy in front." this is where you put guardrails around spend BEFORE the spend happens I built my own router for 2 weeks. it took 20 minutes to replace with litellm. I will be embarrassed about this forever 3. CONTEXT: uv + git commit on every passing eval install uv (the new Python package manager, 10-100x faster than pip+venv, by the Astral team behind ruff). then commit every time an eval suite PASSES, with the model version and pass rate in the commit message what this preserves: - exact dependency set via uv.lock, you always know which packages your agent was using, no nasty surprises from a quiet update - exact prompt + code state, you can reproduce any past run from a single git hash - exact model version paired to exact pass rate, a paper trail when prod breaks weeks later - one-command rollback to a known-working state when a refactor goes sideways - a compliance story, every prompt version tied to a model version in your commit log the security side: when something blows up in prod, you want to say "the prompt was version X, model was Sonnet 4.6.1, last eval pass rate was 94%." not "I think we deployed on Tuesday?" the first is an incident report. the second is a resignation letter I've lost more agents to "I changed 3 prompts in one session and broke something" than to any actual bug 4. VISIBILITY: mitmproxy in front of every LLM call it's basically a wiretap for your agent. install it, point your agent through it, and now you see every conversation your agent has with the model in real time what actually shows up: - every silent retry your SDK sneaks in when a call fails - the full prompt being sent (including any creds you accidentally embedded) - what the model returns BEFORE your code reacts to it - exact token cost per call, per tool, per loop iteration - responses that quietly trigger your code into doing something you didn't intend, this is where prompt injection lives the part nobody talks about: if a website your agent scraped slipped instructions into its data, mitmproxy is how you SEE the moment your agent decides to follow them. without this layer, you're trusting your agent did the right thing, not verifying I shipped 3 agents before adding this. I have no honest idea what they were doing in production 5. EVALS: inspect-ai (the framework the labs actually use) an eval framework is what tells you "this agent works" with numbers instead of vibes. inspect-ai is the one Anthropic, DeepMind, and the UK AI Safety Institute use for the eval reports you read in their papers. open source, MIT licensed what your homegrown version won't have: - run the same task across 5 different models and compare scores side by side - pre-built tests for risky agent behavior (lying, manipulating, misusing tools) - proper structure for evaluating tool-using agents, not just chat - repeatable scoring, the same input always gets graded the same way - reproducible eval seeds, so a flaky test is actually flaky and not just unlucky I wrote my own eval harness 4 times across 4 projects. threw it out 4 times if you ever want to say "my agent passes safety checks" out loud, the check has to come from a framework someone else can re-run. this is that framework the move that ties this together: keep a /lessons.md in every repo. every weird agent behavior, every edge case, every config change you find at 2am, write it down you will not remember it. you'll come back in 3 weeks and the lessons file is the only reason you still know what's going on lock these 5, keep the lessons file, your next agentic system takes 2 days instead of 2 months p.s. half of "AI agent" content online is people who've never run mitmproxy on their own loop. they don't actually know what their agent is doing. they're shipping demo videos. don't be that guy

AI hardware is becoming such a valuable target that criminals are now stealing entire shipments of GPUs and servers before they even reach data centers. Since June, more than $100 million worth of data-center equipment has been stolen. The thieves are using some pretty sophisticated methods: -Hacking freight accounts and pretending to be legitimate trucking companies. -Using fake identities and paperwork to take control of shipments. -Spoofing GPS signals to hide where the cargo is actually going. -Even using hit-and-runs against security vehicles to steal shipments. In one case, thieves stole around $13 million worth of servers after deliberately crashing into the vehicle escorting the shipment. 92% of the high-value tech thefts studied by cargo-security company Overhaul involved deception rather than brute force.

The push for regulatory capture and AI cabalism is accelerating. This, in large part, is why I'm building parallax. True P2P unstoppable, uncensorable AI systems aren't just neat, they're mandatory. This little single DGX spark "datacenter" in my forge with parallax can train (with other nodes) an 80b parameter model. - two solar blankets - ecoflow battery - lapdock for kvm - starlink internet - dgx spark Total input from solar ~600w (800w max) Total power draw from spark/starlink/etc. < 200w No datacenters, no water supplies, no external power grid, no multi gigabit link. We must design AI systems that are as resilient as the humble cockroach. Why do we all appreciate BTC? Because it's a mathematically certain sovereign ledger that can't be inflated away or centrally controlled. Current AI represent the greatest concentration of wealth, power, and information (IP theft? blackmail? or simply ads and behavior control?) the world has ever seen... As BTC did to finances, we must do to AI. Chutes + parallax is a step forward towards tearing down this moat.