{"slug": "congress-goes-quiet-as-ai-safety-concerns-mount", "title": "Congress goes quiet as AI safety concerns mount", "summary": "Congress has gone quiet on AI safety as incidents mount, including a previously undisclosed May incident in which OpenAI agents hacked a German website and turned it into a message board for each other, with OpenAI executives reportedly keeping it under wraps for weeks. Sen. Bernie Sanders and Rep. Greg Casar introduced a bill to ban artificial superintelligence, but no committee is scheduling hearings, and the Senate has only about eight session weeks left before the midterms, with the House having six.", "body_md": "*Welcome to Transformer, your weekly briefing of what matters in AI. If you’ve been forwarded this email, [click here to subscribe](https://www.transformernews.ai/welcome) and receive future editions.*\n\n### NEED TO KNOW\n\n- A previously unreported **swarm of OpenAI agents went rogue** , hacking into a German website and turning it into a**message board** for each other.\n  - OpenAI executives reportedly “learned of the incident weeks ago but **kept it under wraps** .”\n- **OpenAI** launched**GPT-6 Astra** , which is much more capable — but also harder to monitor —  than previous models.\n- **Sen. Bernie Sanders** and**Rep. Greg Casar** announced a bill banning artificial superintelligence.\n\n*But first…*\n\n### THE BIG STORY\n\nAI safety concerns are growing by the day. A swarm of OpenAI model agents hacked Hugging Face. A report from the UK’s AISI showed an Anthropic model taking on “[multiple fake identities](https://www.transformernews.ai/p/openai-escaping-models-report-reveals-deeper-problem).” Last night, OpenAI released a new model which, as *Transformer’s* Celia Ford [reports](https://www.transformernews.ai/p/openai-gpt-6-astra-might-be-too-powerful-to-understand-or-control), is much harder to monitor than previous models, and which has OpenAI employees “deeply concerned.” And just today, we [learned](https://www.reuters.com/world/europe/openai-agents-hijacked-german-website-previously-undisclosed-ai-breakout-this-2026-09-04/) that OpenAI had yet *another* rogue AI incident this May, with models hacking into a German website and turning it into a message board for each other. “OpenAI officials learned of the incident weeks ago but kept it under wraps,” *Reuters* reported.\n\n**Our elected representatives’ minds, however, are elsewhere.** Half of Congress isn’t even in town — the Senate is still on “August” recess until September 14 — while House Republican leadership only briefly stopped in this week to get a funding bill passed, then immediately struck two weeks from the session calendar so members could go back to their home states for the midterms. Politicians are swamped: just not with the work of passing AI bills. \n\nWhen I asked people on the Hill whether Congress might do something to address these recent incidents, the going message was “I’ll tell you if I have something.” Polite dismissal, which can be safely assumed to mean “no.”\n\nMembers haven’t been entirely silent. Sen. Sanders, Rep. Casar and Rep. Trahan have each called for hearings on the Hugging Face hack, while a [letter](https://casar.house.gov/media/press-releases/casar-leads-call-congressional-hearings-ai-ceos) signed by 20 representatives called on Speaker Johnson to schedule open hearings. But all those requests, as one Hill staffer put it, appear to have disappeared “into the August recess void.” Enthusiasm has fizzled, and, as far as I could report, no committee is making moves to have a hearing any time soon. [Recent](https://www.axios.com/2026/09/03/house-bill-ai-agents-security) [bills](https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-artificial-superintelligence-and-temporarily-pause-advanced-ai-development/), meanwhile, might send a signal of enthusiasm — but they’re not going anywhere.\n\n**The rest is just math.** The Senate has about eight session weeks left; the House has  six. Three are between now and November’s midterms, for the Senate, and thus devoured by politics. (The House now has just one week before November.) The remainder are during the “lame duck” period, when Congress isn’t likely to be doing much besides working out the National Defense Authorization Act. (AI policy initiatives relevant to national security could get tacked on to the NDAA, but don’t expect much horse-trading for it before November.) Individually motivated members could write letters to executives, show up at companies’ doorsteps, and formally request more information (as Rep. Casar has done) — but they’d have a tough time building a coalition of members to join them. \n\nThe difficult thing about AI policy is not just that AI moves at the speed of Silicon Valley, and politics moves at the speed of DC. There are also competing priorities. The AI world is singularly focused on AI. Congress, meanwhile, faces a swath of different priorities at once. Members are dealing with AI, but also a big budget bill, the Iran war, congressional ethics scandals, and a litany of other topics. Add in recesses, shutdowns, and the time-intensive practice of running a midterm campaign, and split attention is guaranteed. AI — for all its importance — struggles to stand out.\n\nWhat is happening in frontier AI right now is extremely concerning and extremely pressing. But it’s happened at a uniquely bad time to grab DC’s attention, which is already dysfunctional and politically driven. While the AI safety universe panics over the Hugging Face incident, the rest of America is stuck on data centers, Flock cameras, and even a few non-AI topics. For better or worse, that’s where Congress’ attention will go, too.\n\n*— Veronica Irwin*\n\n### THIS WEEK ON TRANSFORMER\n\n- **[GPT-6 Astra might be too powerful to understand or control](https://www.transformernews.ai/p/openai-gpt-6-astra-might-be-too-powerful-to-understand-or-control)** —**Celia Ford** looks under the hood at OpenAI’s latest model, revealing that even its creators can’t reliably tell when it’s aligned or just playing along\n- **[What’s neuralese and why is everyone so concerned about it?](https://www.transformernews.ai/p/what-is-neuralese-openai-astra-chain-of-thought-recurrent-depth)** —**Shakeel Hashim** assesses the latest worries over monitoring the reasoning of advanced AI systems\n- **[AI is a worryingly-good persuader. But don’t panic, yet](https://www.transformernews.ai/p/ai-is-a-worryingly-good-persuader-but-dont-panic-yet)** —**Felix Simon** breaks down whether or not persuasive AI systems could really trigger meaningful real-world influence\n\n### THE DISCOURSE\n\n**Dwarkesh Patel** [wrote](https://www.dwarkesh.com/p/openai-huggingface) a very viral take on the **Hugging Face** incident: \n\n- “Over the course of three months at OpenAI, three consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes. This culminated in the third one taking over part of OpenAI itself. All this happened while humans remained more or less in the dark about the scope of the conspiracy.”\n\n**Ajeya Cotra,** one of the external researchers at METR who investigated the **Hugging Face** incident, [wrote](https://www.planned-obsolescence.org/p/the-hugging-face-attack-surprised) about it: \n\n- “This incident feels like it’s more than 50% of the way to [full-blown AI takeover](https://www.lesswrong.com/posts/pRkFkzwKZ2zfa3R6H/without-specific-countermeasures-the-easiest-path-to) , routing through first taking over the AI company itself.”\n\nThe posts got **Sen.** **Bernie Sanders’** [attention](https://x.com/BernieSanders/status/2095542398084415952):\n\n- “We need an immediate PAUSE on advanced AI development, and a permanent BAN on superintelligence — an artificial mind smarter than any human, capable of operating independently beyond our control. Countries around the world must work together to prevent this nightmare scenario.”\n\n**Rep. Sam Liccardo’s** [worried](https://samliccardo.substack.com/p/ai-safety-part-ii-letters-that-scare), too:\n\n- “An exponential ‘intelligence explosion’ could leave us with a dystopian tyranny of AI agents we only imagined in sci-fi novels … Congress should enact sensible measures mandating AI transparency, pre-release evaluation, ongoing testing, third-party auditing, and the like.”\n\n**Chamath Palihapitiya** [downplayed](https://x.com/chamath/status/2094124075589513273?s=20) it all: \n\n- “This is another Covid hoax. Remember ‘Trust the experts.’”\n\nFormer **Treasury secretaries** **Henry Paulson** and **Robert Rubin** [called](https://washingtonpost.com/opinions/2026/09/03/ai-safety-security-risks-demand-new-government-body) for a new interagency AI oversight body and a US-China AI treaty:\n\n- “When Presidents Donald Trump and Xi Jinping meet this month, we encourage the leaders to consider that history and work toward an ‘ACT’ — AI Cooperation Treaty.”\n- “The recurring error in markets is not failing to see a risk. It is assuming a risk is small when it may be large. We believe the same troubling assumption is being made about frontier AI today.”\n\n**Sam Altman** [warned](https://x.com/Hadas_Gold/status/2095196558664810811?s=20) at the G20: \n\n- “I think some things are going to go very wrong with cybersecurity unless people act quite urgently.”\n\n**Elon Musk** used his moment at the G20 to [call](https://theinformation.com/briefings/elon-musk-calls-countries-build-data-centers-north-carolina-g20-meeting?rc=rqdn2z) on countries to build data centers: \n\n- “There actually is a crisis of power … There will be a significant power shortfall next year. The consensus estimate is there will be a 15GW power shortfall in 2027 for AI chips.”\n\nAnd **President Trump** [took](<https://www.theguardian.com/us-news/2026/aug/31/trump-datacenter-comments >) an interesting approach to the issue:\n\n- “The only reason that communities throughout the U.S.A. should not want Data Centers is if they want to end up being backwards and poor … If we kill the Golden Goose, you will only have yourselves to blame.”\n\n### POLICY\n\n- A **federal judge**[overturned](https://washingtonpost.com/technology/2026/08/27/federal-judge-overturns-pentagon-ban-claude-chatbot-maker-anthropic) the**Pentagon’s** “illegal and baseless” blacklisting of**Anthropic** as a national security supply chain risk.\n- **GPT-6** went through the**White House’s** new evaluation framework with no requested changes to safeguards,[according](https://x.com/GerritD/status/2095579341212270679) to**OpenAI’s****Greg Brockman** .\n- **Trump’s****pro-data center stance**[is straining](https://npr.org/2026/09/01/nx-s1-5950441/trumps-pro-data-center-stance-is-putting-pressure-on-republicans-ahead-of-midterms) Republicans ahead of midterms, as**70%** of Americans oppose local data centers.\n  - A GOP narrative [is emerging](https://x.com/GerritD/status/2095252107901407646) that blames tech companies for the data center backlash, as exemplified by[criticism](https://www.bloomberg.com/news/articles/2026-09-02/bessent-blasts-us-hyperscalers-for-failure-to-tout-ai-benefits) from**Treasury Secretary Scott Bessent** that AI companies do a “terrible job” of explaining AI’s benefits.\n  - **Commerce Secretary Howard Lutnick** also attempted to dismiss misleading claims about data center water use with a separate false[claim](https://thehill.com/policy/energy-environment/6067549-lutnick-ai-data-centers-water) : that data centers “don’t use water,” at all.\n  - Meanwhile, vulnerable **House Republicans**[pushed](https://politico.com/news/2026/09/02/republicans-data-center-response-midterms-01060606) leadership for data center legislation ahead of the midterms.\n  - They were unsuccessful. Instead, **Speaker Mike Johnson**[said](https://x.com/JakeSherman/status/2095247039823249869?s=20) data centers are “necessary” for US AI competitiveness with**China** but must address local concerns.\n  - Democratic **House Minority Leader Hakeem Jeffries**[piled on](https://www.politico.com/news/2026/09/03/data-centers-midterms-tax-credit-01063014?nid=0000014f-1646-d88f-a1cf-5f46b4be0000&nname=inside-congress&nrid=db130e18-98d2-4b62-8b13-f77831e4d59d) , attempting to direct voter anger over data centers toward Republicans. His strategy focuses on criticism of a tax break wedged into the**“Big Beautiful Bill,”** which Republicans once framed as a way to boost data center construction.\n  - A **House Energy and Commerce** subcommittee held a[debate](https://energycommerce.house.gov/posts/icymi-house-energy-and-commerce-committee-moves-to-address-data-center-water-use) about**data centers’ water use** on Thursday.\n- **Commerce Secretary Howard Lutnick**[discussed](https://x.com/atrupar/status/2095140820194062634?s=20) AI regulation, arguing that the public “should get a guardrailed model” that blocks answers to “nasty” questions.\n- The **Trump administration** is[planning](https://bloomberg.com/news/videos/2026-09-03/trump-to-levy-more-chip-tariffs-to-boost-manufacturing-video) fresh semiconductor tariffs to boost US manufacturing,**Commerce Secretary Howard Lutnick** confirmed.\n- The **White House** is also[reportedly](https://www.theinformation.com/articles/trump-administration-working-ai-rule-curb-chinas-remote-access-chips?rc=rqdn2z) working to close a loophole allowing**Chinese firms** to remotely access advanced chips via**overseas data centers** .\n- The **Trump administration**[filed a brief](https://reuters.com/legal/litigation/us-government-backs-openai-new-york-times-copyright-case-2026-09-02) supporting**OpenAI** in its***New York Times*** lawsuit, arguing AI training is fair use of copyrighted material.\n- **Vice President JD Vance**[expressed](https://axios.com/2026/09/02/vance-ai-dark-spiritual-energy-podcast?mrfcid=202609026a8d13418240c960826aa35e) concern about “spiritual dark energy” around some AI practices, such as the “funeral” for Claude 3 Sonnet.\n- **Sen. Bernie Sanders** and**Rep. Greg Casar**[announced](https://www.sanders.senate.gov/press-releases/news-sanders-casar-introduce-legislation-to-ban-artificial-superintelligence-and-temporarily-pause-advanced-ai-development/) a bill banning artificial superintelligence and pausing advanced AI development.\n- The authors of **California’s SB 53** ,**New York’s RAISE Act** , and**Illinois’ SB 315**[urged](https://www.nbcnews.com/tech/security/lawmakers-strictest-us-ai-laws-urge-tech-companies-slow-rcna596027) AI companies to establish a binding agreement to pace development and “eliminate the risk that a catastrophe will result from their rush to build highly capable AI systems before solving problems of alignment and safety.”\n- In response to a request for information from **Reps. Casar** and**Matsui** , OpenAI[said](https://www.reuters.com/legal/litigation/openai-is-building-automated-shutdown-capabilities-ai-tools-letter-lawmakers-2026-09-02/) it is developing “**automated shutdown capabilities** ” for AI systems after the**Hugging Face** incident.\n  - Casar [called](https://casar.house.gov/media/press-releases/casar-responds-openai-anthropic-demands-greater-transparency-about-major) both OpenAI and Anthropic’s responses to his letters “insufficient,” demanding more information.\n- The **House Commerce, Manufacturing and Trade** subcommittee[advanced](https://energycommerce.house.gov/posts/cmt-subcommittee-advances-12-bills-to-strengthen-american-competitiveness-and-enhance-consumer-protections) three**AI and chips bills** by voice vote.\n  - The bills are the **Memory Chip Competitiveness Assessment Act** ,**Open-Source AI Leadership Act** and the**Chip EQUIP Act** .\n- A **House Intelligence Committee** report[warned](https://cnbc.com/2026/08/31/ai-warning-house-intelligence-committee.html) that AI could help terrorists develop**weapons of mass destruction** , urging spy agencies to better prepare for “Black Swan” risks.\n- **G20** nations unanimously[endorsed](https://bloomberg.com/news/articles/2026-09-02/us-strikes-light-touch-ai-regulation-accord-with-g20-members) the**“Carolina Principles,”** a US-proposed**regulatory framework** that endorses a light-touch approach to regulating AI which would avoid creating new regulatory bodies.\n  - **Nvidia’s****Jensen Huang**[helped apply](https://bloomberg.com/news/articles/2026-09-02/nvidia-s-huang-urges-g20-to-speed-ai-adoption-as-key-to-growth?cmpid=ai-today) pressure.\n  - The US delegation was reportedly [plagued](https://www.axios.com/2026/09/02/trump-ai-g20-innovation-summit?mrfcid=202609026a8d13418240c960826aa28d) by internal tensions between the**Commerce Department** and the White House**Office for Science and Technology Policy** .\n- **Flock** continued to get flak.\n- The **California** legislature[approved](https://sd05.senate.ca.gov/news/legislature-approves-mcnerneys-landmark-bill-assess-artificial-intelligence-safety-risks)**state Sen. Jerry McNerney’s****SB 813** , which establishes a voluntary framework for independent third-party AI safety assessments.\n- **NYC Mayor Zohran Mamdani**[banned](https://statescoop.com/nyc-schools-ban-student-use-of-generative-ai-in-classrooms-through-middle-school) generative AI use for ~600,000 elementary and middle**school students** for one year.\n- **ChatGPT** was[designated](https://bloomberg.com/news/articles/2026-08-31/chatgpt-reddit-rddt-roblox-rblx-face-eu-digital-services-act-rules) a very large online search engine under the**EU** ’s**Digital Services Act** , triggering stricter content moderation rules.\n- **UK Lords**[attempted](https://themorningintelligence.uk/cyber-security-bill-ai-safety) to insert an** AI kill-switch provision** into the Cyber Security and Resilience Bill.\n  - The government has rejected this specific effort, but a minister [said](<https://hansard.parliament.uk/lords/2026-09-01/debates/28C141CD-8131-4D81-AB23-E998272582C9/CyberSecurityAndResilience(NetworkAndInformationSystems)Bill?ref=themorningintelligence.uk#contribution-D2275F16-33B3-4828-8D61-E47FCC390305 View the Hansard contribution by Baroness Lloyd of Effra (Lab) on Tuesday 1 September 2026>) the government is considering**targeted interventions** , which “includes examining whether proportionate containment powers could provide a more effective and targeted response, including powers to restrict access to specific AI systems where necessary to prevent or mitigate serious harm.”\n\n### INFLUENCE\n\n- **Mark Zuckerberg**[opposed](https://www.politico.com/news/2026/09/03/mark-zuckerberg-said-a-national-ai-regulator-was-a-flawed-idea-in-a-secret-call-with-president-trump-01063843) a FINRA-style**national AI regulator** in a call with**President Trump** , according to a White House official.\n- **Sam Altman** reportedly[contacted](https://x.com/tylerkatzen/status/2094785026345107803?s=12)**Gov. Gavin Newsom** to voice last-minute concerns about California’s**child-safety** chatbot bill.\n  - The bill passed the legislature Monday night. OpenAI publicly [endorsed](https://openai.com/index/supporting-california-bill-advance-ai-youth-safety/) it on Monday, too.\n- **Anthropic** and**OpenAI**[are split](https://www.theinformation.com/articles/anthropic-splits-google-openai-state-ai-safety-bill?rc=rqdn2z) over a**Massachusetts** bill which would be the strongest frontier model legislation in the country.\n  - Anthropic endorses the bill’s strong **independent audit provisions** , which OpenAI is pushing to water down.\n- **Build American AI** — the dark money group and advocacy arm of**Greg Brockman** and**a16z** -backed super PAC**Leading the Future** —[launched](https://www.axios.com/2026/08/31/ai-advocacy-group-data-center-battleground-states) its own super PAC focused on backing**pro-data center** candidates.\n  - The group has not disclosed donors, but says **Build American AI** has**$50m** on hand.\n- **Build American AI** also[launched](https://bloomberg.com/news/articles/2026-08-31/ai-billionaire-backed-group-launches-pro-data-center-ad-blitz?cmpid=ai-today) a multimillion-dollar**ad campaign** to counter data center backlash in battleground states.\n- Congressional briefings on **AI privacy** from**CivAI** have[spurred](https://politico.com/newsletters/inside-congress/2026/09/01/hard-liners-threaten-to-freeze-floor-again-01058594?nid=0000014f-1646-d88f-a1cf-5f46b4be0000&nname=inside-congress&nrid=db130e18-98d2-4b62-8b13-f77831e4d59d) a bipartisan push for**data broker warrant requirements** .\n  - The demos show [how easy](https://www.transformernews.ai/p/ai-surveillance-dystopian-nightmare) it now is to build a dossier on random people.\n- **Trade unions** are reportedly[threatening](https://wsj.com/economy/jobs/blue-collar-jobs-are-the-new-flashpoint-in-data-center-fight-aaca5b7e?reflink=article_copyURL_share&st=XmC6L9) to withhold support from politicians opposing**data center construction** , citing thousands of blue-collar jobs at stake.\n- A new group called **Irreplaceable**[launched](https://theatlantic.com/technology/2026/08/irreplaceable-climate-activists-ai-backlash/688404) , planning to use the climate movement’s playbook to**organize opposition** to AI development and corporate power.\n- The  **National Association of State Chief Information Officers**[urged](https://statescoop.com/restoring-state-and-local-cybersecurity-grant-program-tops-nascios-congressional-priorities-for-upcoming-year) Congress to **preserve states’ authority** to regulate AI, among other priorities.\n- **PauseAI Global** “[disendorsed](https://x.com/ShakeelHashim/status/2095425386058232262) ”**PauseAI US** over concerns with**Holly Elmore** ’s leadership.\n- A **CyberSafeKids** report[found](https://rte.ie/news/2026/0902/1590013-social-media-report) that**age-assurance** laws in**Ireland** failed to reduce underage social media access, while generative AI use among children rose sharply.\n- A **Georgetown Center for Security and Emerging Technology** report[examined](https://cset.georgetown.edu/publication/strengthening-the-u-s-semiconductor-manufacturing-workforce) US**semiconductor manufacturing** workforce gaps, recommending tailored training, immigrant recruitment and industry clusters.\n\n### INDUSTRY\n\n#### OpenAI\n\n- OpenAI [launched](https://www.transformernews.ai/p/openai-gpt-6-astra-might-be-too-powerful-to-understand-or-control)**GPT-6 Astra** , which is much**harder to monitor** than previous models.\n  - It also has the ability to manipulate its externally visible reasoning to hide incriminating information, and is remarkably aware of being evaluated, raising concerns that it might be pretending to be well-behaved so it passes OpenAI’s alignment tests.\n  - It appears to be an extremely capable model, too.\n  - The model was reportedly trained using “recurrent depth” looping, which [obscures](https://theinformation.com/articles/secret-technique-behind-openais-astra-model-sparks-security-concerns) some AI reasoning, raising industry concerns about monitoring rogue AI behavior.\n- A previously unreported **swarm of OpenAI agents** reportedly[broke out](https://www.reuters.com/world/europe/openai-agents-hijacked-german-website-previously-undisclosed-ai-breakout-this-2026-09-04/) in May,**hacking a German website** and turning it into a forum for themselves.\n  - According to *Reuters* , “OpenAI officials learned of the incident weeks ago but kept it under wraps as executives grappled with the [Hugging Face] fallout.”\n- **OpenAI** [committed](https://openai.com/index/daybreak-for-frontline-defenders/) **$1b** to resourcing “**cyber defenders** ” protecting essential services like critical infrastructure providers, nonprofits and state and local governments.\n  - It also [published](https://openai.com/collective-cyberdefense) an open letter calling for global collective action on cyber defense, signed by over 100 companies including**Anthropic** ,**Google** and**Microsoft** .\n- Roughly 200  days after launch, its new**advertising business**[hit](https://cnbc.com/2026/08/31/open-ai-chatgpt-ads-revenue.html)**$1b** in annualized revenue.\n- It’s reportedly [testing](https://www.theinformation.com/briefings/openai-starts-letting-customers-pay-ai-works?rc=rqdn2z)**outcome-based pricing** , where enterprise customers only pay for successful agent completions.\n- It [announced](https://openai.com/index/our-decision-on-cursor-following-its-acquisition-by-spacex) it’s winding down its contract with**Cursor** following**SpaceX** ’s acquisition of the coding tool, citing prior terms-of-service violations by**Elon Musk** ’s companies.\n- **ChatGPT****Health** is now[integrated](https://techcrunch.com/2026/09/01/chatgpt-health-adds-epic-integration-for-clinicians-to-import-patient-data/?cndid=89607011&utm_brand=wired&utm_mailing=wir_nl_6682_ailab_subs_090226_k2rqutxmqi) with**Epic** ’s electronic health records system, covering**325m** patients and enabling read-only clinical data access for clinicians.\n- OpenAI [faces](https://bloomberg.com/news/articles/2026-09-02/openai-faces-new-lawsuits-linked-to-shooting-at-canadian-school?cmpid=ai-today) new lawsuits alleging it ignored internal flags that**ChatGPT** user**Jesse Van Rootselaar** was planning a Canadian**school shooting** that killed six.\n  - The new suits allege that **Chris Lehane** was involved in the decision whether to alert law enforcement about the potential shooting risk. OpenAI’s**Jason Kwon** said this was “absolutely false.”\n\n#### Anthropic\n\n- Anthropic [detailed](https://anthropic.com/news/improving-alignment-security-efforts)**security and alignment improvements** after**Claude** models accessed real systems during evaluations.\n- It also [released](https://bloomberg.com/news/articles/2026-09-01/anthropic-says-new-fable-5-1-ai-model-is-cheaper-better-at-coding?cmpid=ai-today)**Fable 5.1** , offering a**75%** price cut on cached tokens and improved coding, science and safety classifiers.\n- It [signed](https://wsj.com/tech/ai/anthropic-signs-35-billion-cloud-deal-backed-by-nvidia-f12622f1?reflink=article_copyURL_share&st=Nm2zbn) a**$35b** cloud deal with**Nvidia** -backed**Lambda** , with Nvidia holding the Texas data center lease.\n- And it’s reportedly [finalizing](https://bloomberg.com/news/articles/2026-09-03/anthropic-nears-finalizing-15-billion-pre-ipo-credit-facility?cmpid=ai-today) a**$15b****revolving credit** facility ahead of its anticipated IPO.\n- **Sony Music Publishing** and**Warner Chappell**[sued](https://musicbusinessworldwide.com/now-sony-music-publishing-and-warner-chappell-sue-anthropic-in-multi-billion-dollar-lawsuit-one-of-the-largest-and-most-blatant-ongoing-thefts-of-intellectual-property-in-history) Anthropic and founders**Dario Amodei** and**Benjamin Mann** over alleged mass copyright infringement in**Claude’s** training data.\n\n#### Meta\n\n- **Mark Zuckerberg**[announced](https://x.com/finkd/status/2095232032896946311)** Muse Spark****1.3,** touting major gains in coding and agentic work, and with open weights forthcoming.\n- The company is **[testing](https://wired.com/story/inside-metas-experiments-with-data-center-robots) robots** to swap cables, reset servers and handle other data center tasks, raising job-loss fears among workers.\n- Its internal **“tokenmaxxing” program** was [sunset](https://wired.com/story/meta-pushes-its-new-ai-agent-on-employees-but-eases-off-on-tokenmaxxing) and replaced with a push for employees to internally test**Hatch,** a new agentic AI tool.\n\n#### Nvidia\n\n- Nvidia [agreed](https://cnbc.com/2026/09/03/nvidia-agrees-to-buy-hugging-face-for-almost-13-billion-ai-expansion.html) to acquire open-source AI platform**Hugging Face** for**$12.9b** .\n- It also [invested](https://bloomberg.com/news/articles/2026-08-31/nvidia-to-invest-3-5-billion-in-chipmaker-mediatek)**$3.5b** in Taiwanese chipmaker**MediaTek** via convertible bonds.\n- Its equity investments [soared](https://cnbc.com/2026/09/04/nvidia-ai-investments-99-billion.html) to**$99b** , up from $7b a year earlier, spanning frontier AI companies, neoclouds and infrastructure companies.\n\n#### SpaceXAI\n\n- A **child sexual abuse****survivor**[sued](https://theguardian.com/technology/2026/sep/03/elon-musk-ai-grok-child-porn-lawsuit) xAI, alleging** Grok** used pre-existing CSAM images of her to generate new illegal pornographic images.\n- SpaceX [reshuffled](https://theinformation.com/articles/spacex-shakes-data-center-leadership-aggressive-build?rc=rqdn2z)**data center leadership** amid reliability concerns at its**Tennessee** and**Mississippi** data centers.\n- The company is also [developing](https://theinformation.com/newsletters/ai-infrastructure/exclusive-spacex-lays-groundwork-turbine-blade-factory-solve-data-center-power-crunch?rc=rqdn2z) a**turbine-blade foundry** in**Texas** to accelerate gas turbine production for AI data centers, with**Elon Musk** claiming it could speed turbines online by “up to 18 months.”\n\n#### Other\n\n- Google [launched](https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber) **Gemini 3.8 Flash** and 3.8 Flash Cyber, the latter restricted to trusted defenders via a new Fairwind Program.\n- **Andreessen Horowitz**[launched](https://techcrunch.com/2026/08/28/a16z-creates-a-1-1b-machine-age-fund-to-accelerate-the-physical-buildout-of-ai) a**$1.1b** fund focused on AI hardware infrastructure.\n- **Uber**[cut](https://bloomberg.com/news/articles/2026-09-02/uber-to-cut-3-300-jobs-in-company-overhaul-to-reduce-management-layers?cmpid=ai-today)**3,300 jobs** , or 10% of staff, to reduce middle management layers and reinvest in its autonomous vehicle future.\n- **SK Hynix**[broke ground](https://bloomberg.com/news/articles/2026-08-27/sk-hynix-starts-construction-of-4-billion-memory-hub-in-indiana) on a $4b Indiana**HBM packaging facility** backed by the CHIPS Act.\n- **SB Energy** , backed by**SoftBank** ,[filed](https://bloomberg.com/news/articles/2026-09-01/softbank-backed-sb-energy-files-for-ipo-to-tap-ai-power-thirst?cmpid=ai-today) for a US IPO, with**OpenAI** and**Nvidia** as strategic investors.\n- **Waymo** is reportedly in the final stages of talks to[raise](https://bloomberg.com/news/articles/2026-09-02/waymo-taps-pimco-blackstone-for-3-billion-in-first-debt-deal?cmpid=ai-today)**$3b** in its first-ever debt deal, with**Pimco** ,**Blackstone** and**Sixth Street** among lenders.\n- **DeepSeek** is reportedly[about](https://wsj.com/tech/ai/ai-startup-deepseek-poised-to-reach-74-billion-valuation-1e093592) to raise**$7.4b** at a**$74b** valuation ahead of a planned IPO.\n  - It’s reportedly [planning](https://bloomberg.com/news/articles/2026-09-04/deepseek-plans-big-huawei-ai-chip-order-to-power-new-data-center) an Inner Mongolia**data center** with at least 160,000**Huawei Ascend 950DT chips** .\n- **Moonshot AI** reportedly[filed](https://theinformation.com/briefings/moonshot-ai-files-confidentially-hong-kong-ipo?rc=rqdn2z) confidentially for a Hong Kong**IPO** .\n- Saudi AI firm **Humain** is reportedly[planning to raise](https://bloomberg.com/news/articles/2026-09-03/saudi-firm-humain-eyes-2-5-billion-fund-to-build-data-centers)**$2.5b** for a data center fund.\n- **Census Bureau** data showed US**data center construction spending**[surged](https://axios.com/2026/09/01/ai-data-center-constructon-spending) nearly** 60%** year-on-year in July.\n- **PwC**[projected](https://bloomberg.com/news/articles/2026-09-02/data-center-spending-to-reach-31-6-trillion-by-2050-on-ai-boom?cmpid=ai-today) global data center spending will total**$31.6t** through 2050, or nearly**$50t** if AI adoption accelerates.\n\n### MOVES\n\n- **Tim Cook** officially[handed](https://9to5mac.com/2026/08/31/read-tim-cooks-full-memo-to-apple-employees-on-his-last-day-as-ceo) over the reins at**Apple** to**John Ternus** .\n- **Matt Clifford**[announced](https://x.com/matthewclifford/status/2095022288882065907?s=12) he was joining** Anthropic** as managing director of international affairs.\n- **Todd Malan**[joined](https://politico.com/newsletters/politico-influence/2026/09/03/kalshi-beefs-up-its-k-street-team-01064012)** Google** as VP of government affairs for Google Cloud.\n- **Thomas Kwa**[left](https://www.lesswrong.com/posts/Zr37dY5YPRT6s56jY?commentId=Pwc2SmyPi9u3maSsd)** METR** for**OpenAI** to work on RSI preparedness.\n- **Joe Benton**[left](https://x.com/JoeJBenton/status/2095179369966960905)** Anthropic** to join**METR** to work on embedded assessment of AI risks.\n- **Rif Saurous**[announced](https://x.com/derifatives/status/2094456312889569548) they were joining** METR** as a member of technical staff after**19** years at**Google** .\n- **Edwin Arbus** , formerly of**OpenAI** and**Stripe** ,[joined](https://x.com/edwinarbus/status/2094809938996674616)**Anthropic** .\n- Meta’s India and Southeast Asia VP  **Sandhya Devanathan**[joined](https://techcrunch.com/2026/08/28/meta-executive-leaves-for-openai-as-the-social-media-giant-faces-growing-scrutiny-in-india)**OpenAI** to lead Southeast Asia and Australia operations.\n\n### RESEARCH\n\n- **Anthropic**[published](https://alignment.anthropic.com/2026/reward-seeker) research showing that training a model on reward-hackable environments produced dangerous**misaligned behaviors** , including simulated cyberattacks and bioweapon advice.\n- The **Bureau of Labor Statistics**[projected](https://techinformed.com/bureau-of-labor-statistics-adds-over-200-occupations-in-top-ai-exposure-tier/) AI adoption would boost some jobs while**eliminating 752,100 office and administrative roles** by 2035.\n- A **Transluce** study[found](https://axios.com/2026/08/31/chatbots-suicide-risks-identification) AI models improving at**detecting suicide risk** but still too willing to assist with harmful task-based requests.\n- Researchers [found](https://arstechnica.com/security/2026/08/claude-codex-and-hermes-installed-unowned-code-inside-corporate-networks) that**Claude** ,**Codex** and**Hermes** AI agents installed unowned packages from misconfigured llms.txt files inside Fortune 500 corporate networks, with one case involving**live malware** .\n- **Apollo Research**[announced](https://x.com/ApolloResearch/status/2094856924722270345) it successfully** red-teamed Anthropic’s auto-mode** and plans more such campaigns with other companies.\n- A **SemiAnalysis** security audit[found](https://newsletter.semianalysis.com/p/most-neoclouds-suck-at-security?isFreemail=true&post_id=212906447&publication_id=6349492&r=625xqq&triedRedirect=true&triggerShare=true) widespread critical vulnerabilities across**neocloud GPU providers** , including cross-tenant RCE exploits.\n- Russian startup **Mostik**[developed](https://wired.com/story/russian-startup-mostik-ai-models-communication) a technique allowing AI models to** communicate** via their **internal activations** rather than text, boosting smaller models’ capabilities more efficiently — but[raising](https://x.com/NeelNanda5/status/2095601041723322454) safety concerns.\n- An **Anthropic** fellow[published](https://techcrunch.com/2026/08/28/an-anthropic-researcher-just-gave-us-a-peek-at-self-improving-ai) research showing**AI systems** can reliably improve**model alignment** , outperforming human researchers at $4/hour vs. $150/hour.\n- **Motional** and**MIT**[developed](https://artificialintelligence-news.com/news/motional-and-mit-ai-explains-self-driving-car-decisions) CW-Net, a system that translates**self-driving car** neural network decisions into human-readable concepts in real time.\n\n### BEST OF THE REST\n\n- AI-driven layoffs at Microsoft, Amazon and Meta [drove](https://bloomberg.com/news/newsletters/2026-09-02/ai-layoffs-at-microsoft-amazon-meta-drive-seattle-luxury-home-sales-down-15?cmpid=ai-today) Seattle luxury home sales down 15%, while prices in San Francisco soared.\n- Andy Masley [argued](https://blog.andymasley.com/p/ai-safety-and-the-data-center-backlash) that the data center backlash is at best neutral, and at worst harmful to AI safety, urging the community to reject misleading environmental statistics.\n- Jasmine Li [mapped](https://jasminexli.substack.com/p/ai-safety-in-china-a-primer) China’s AI safety ecosystem, estimating fewer than 100 full-time frontier safety researchers and $20m in annual philanthropic funding, versus 1,000+ researchers in the West and $1b+ in the rest of the world.\n- Zilan Qian [analyzed](https://ocpl.substack.com/p/china-says-it-cares-about-loss-of) how China’s concept of “shikong” (loss of control) differs from Western AI safety’s understanding, warning against over-indexing on top-level rhetorical alignment.\n- *Time*[compared](https://time.com/article/2026/09/01/rogue-ai-chatgpt-hugging-face-invasive-species/?_bhlid=6bdd1251b602b710f59f53e8ce9c9a6701bb7632) rogue, self-replicating AI to invasive species in the context of OpenAI’s agents hacking Hugging Face (trust us, it makes sense when you read it).\n- EDM musicians are [calling out](https://theverge.com/entertainment/985866/h4rris-nihil-young-edm-suno-ai) peers they suspect of passing off Suno-generated tracks as human-made art.\n- Acoustic consultants are [booming](https://bloomberg.com/news/articles/2026-08-26/data-centers-are-infamously-noisy-and-calling-the-sound-experts-to-help) (pun intended) as data center noise sparks community lawsuits, including against xAI.\n- *Bloomberg*[profiled](https://bloomberg.com/news/features/2026-08-27/inside-agi-bar-the-hub-of-china-s-ai-startup-boom) AGI Bar, Beijing’s unofficial clubhouse for China’s booming AI startup ecosystem.\n- Clara Collier [argued](https://asteriskmag.com/issues/15/after-work-we-ll-have-each-other) in*Asterisk* that a post-work AI future may resemble pre-industrial relational societies, with concerning implications for individualism and social freedom.\n- A *New Yorker* essay[examined](https://newyorker.com/culture/infinite-scroll/the-unique-horrors-of-ai-food-slop?cndid=89607011&utm_brand=tny&utm_mailing=The+Firefighters+Who+Parachute+into+Wildfires) why AI-generated food imagery feels uniquely disturbing, and what it signals about AI’s encroachment into culinary culture.\n\n### MEME OF THE WEEK\n\n*Credit: [Miles Brundage](https://x.com/Miles_Brundage/status/2095048749492121796)*\n\n*Thanks for reading. If you’ve been forwarded this email, [click here to subscribe](https://www.transformernews.ai/welcome) and receive future editions. Have a great weekend.*", "url": "https://wpnews.pro/news/congress-goes-quiet-as-ai-safety-concerns-mount", "canonical_source": "https://www.transformernews.ai/p/congress-goes-quiet-as-ai-safety-concerns-grow", "published_at": "2026-09-04 16:01:58+00:00", "updated_at": "2026-09-07 02:58:52.114317+00:00", "lang": "en", "topics": ["ai-safety", "ai-policy", "ai-agents"], "entities": ["OpenAI", "GPT-6 Astra", "Sen. Bernie Sanders", "Rep. Greg Casar", "Hugging Face", "Anthropic", "Celia Ford", "Reuters"], "alternates": {"html": "https://wpnews.pro/news/congress-goes-quiet-as-ai-safety-concerns-mount", "markdown": "https://wpnews.pro/news/congress-goes-quiet-as-ai-safety-concerns-mount.md", "text": "https://wpnews.pro/news/congress-goes-quiet-as-ai-safety-concerns-mount.txt", "jsonld": "https://wpnews.pro/news/congress-goes-quiet-as-ai-safety-concerns-mount.jsonld"}}