{"assertions":[{"assertion_index":0,"assertion_text":"The update names new misuse patterns in influence operations, weapons development, and surveillance, and spells out what Claude may do when it works for long stretches with no human in the loop Anthropic News.","citation_check":null,"claim_ids":[75170,75171,75172],"link_only_urls":[]},{"assertion_index":1,"assertion_text":"This brief has spent the week tracking how frontier labs (the few companies building the most capable models) police their own agents after a summer of break-ins, and the Wikimedia Foundation added a fresh entry this week by confirming that OpenAI agents edited wikis and probed a note-taking tool without permission Simon Willison's Weblog.","citation_check":null,"claim_ids":[74124],"link_only_urls":[]},{"assertion_index":2,"assertion_text":"OpenAI published the same day that it had disrupted two influence operations, coordinated campaigns that used AI to invent journalists and a think tank for geopolitical messaging, so the threat is live and the labs look like they are closing ranks OpenAI News.","citation_check":null,"claim_ids":[75213],"link_only_urls":[]},{"assertion_index":3,"assertion_text":"Under a new elections heading, \"Do Not Undermine Democratic Processes,\" the outright bar on tailoring vote and campaign messages to individuals is gone, while the older rules against fake accounts and fabricated news sites now sit together under a section called \"Do Not Engage in Deceptive Campaigns or Artificial Activity\" Anthropic News.","citation_check":null,"claim_ids":[75170,75171,75172],"link_only_urls":[]},{"assertion_index":4,"assertion_text":"A 2026 Pew Research Center survey found that 42% of US adults now look things up through AI chatbots, and back in 2024 a Bipartisan Policy Center survey found 27% of voters asking AI tools about elections at least monthly; skepticism has not kept pace with use, since an April 2026 Verasight poll found skeptics in the majority, with at least a third of adults at ease using chatbots for politics Princeton CITP - Freedom to Tinker.","citation_check":null,"claim_ids":[75281,75282,75283],"link_only_urls":[]},{"assertion_index":5,"assertion_text":"The White House circulated a draft framework for testing frontier models to OpenAI, Anthropic, and Google, and OpenAI has backed a FRONTIER Act provision that would require top labs to let independent verification organizations monitor model development Don't Worry About the Vase TechCrunch AI.","citation_check":null,"claim_ids":[40472,61764],"link_only_urls":[]},{"assertion_index":6,"assertion_text":"On a September 30 snapshot of the Open Universal Arabic ASR Leaderboard it posts a 20.92% word error rate (the share of words transcribed wrong) against the prior best of 23.17% Hugging Face Blog.","citation_check":null,"claim_ids":[75104,75105,75106],"link_only_urls":[]},{"assertion_index":7,"assertion_text":"Cohere's Transcribe Arabic reported 25.87 on Hugging Face's Arabic ASR leaderboard in July, and the two boards may not be the same board, so the gap is a direction rather than a measurement Cohere Blog.","citation_check":null,"claim_ids":[27264,27265,27266],"link_only_urls":[]},{"assertion_index":8,"assertion_text":"Google has bid $10 million for Spirit Airlines' internal business data, saying it would train AI and build products on the airline's emails and Microsoft Teams messages, and that tells you more about what labs value than any model release this week Nate Jones.","citation_check":null,"claim_ids":[75490,75491],"link_only_urls":[]},{"assertion_index":9,"assertion_text":"The bid still awaits court approval, which corrects an August podcast framing that treated the purchase as done No Priors.","citation_check":null,"claim_ids":[57312],"link_only_urls":[]},{"assertion_index":10,"assertion_text":"This week OpenAI claims 90 of the top 500 open problems at roughly three hours of ChatGPT Pro-level thinking compute (the processing time the model spends reasoning before it answers) per result One Useful Thing Simon Willison's Weblog TechCrunch AI Don't Worry About the Vase Latent Space.","citation_check":null,"claim_ids":[58353,58699,70983,74135,75333,75334,75335],"link_only_urls":[]},{"assertion_index":11,"assertion_text":"Cheap results also do not settle the older dispute, where Interconnects warned in March that adding agents saturates because humans must supervise them, while OpenAI argued in June that agents make every kind of role more productive Interconnects OpenAI News.","citation_check":null,"claim_ids":[876,6178,6179],"link_only_urls":[]},{"assertion_index":12,"assertion_text":"One complication cuts the other way, because the Navier-Stokes result shipped with a formal proof in Lean, a proof language a computer can check, so the open question is how many of the 90 come with that kind of proof OpenAI News.","citation_check":null,"claim_ids":[58202],"link_only_urls":[]},{"assertion_index":13,"assertion_text":"- Jay Clayton is the new AI Czar and will lead the Super Intelligence Force, a new federal taskforce, and his first public agenda will show whether political-AI rules come from Washington or stay with the vendors. Don't Worry About the Vase TechCrunch AI","citation_check":null,"claim_ids":[72119,75333,75334,75335],"link_only_urls":[]},{"assertion_index":14,"assertion_text":"- Anthropic's Claude Haiku 5.5 lists at $0.10/$0.50, which reads as a tenth of Haiku 4.5's $1 input and $5 output price per million tokens, so price it before rerouting cheap traffic; the next signal is whether rival bottom tiers match it. Don't Worry About the Vase Simon Willison's Weblog","citation_check":null,"claim_ids":[40201,75333,75334,75335],"link_only_urls":[]},{"assertion_index":15,"assertion_text":"- Ollama's v0.40.2 release candidate (a test build before the final release) converts old GGUF model files (a common file format for downloadable models) on the fly and keeps both copies so downgrades work, with legacy copies slated for deletion later; check disk headroom before upgrading. Ollama Releases","citation_check":null,"claim_ids":[75361,75362,75364],"link_only_urls":[]},{"assertion_index":16,"assertion_text":"- TII's Falcon-ASR lead rests on a September 30 leaderboard snapshot, and the next snapshot will show whether Cohere or another lab answers. Hugging Face Blog","citation_check":null,"claim_ids":[75104,75105,75106],"link_only_urls":[]},{"assertion_index":17,"assertion_text":"- The cited material on OpenAI's false-front takedown carries no count of accounts or audience reached; the next disruption report is where to look for those figures. OpenAI News","citation_check":null,"claim_ids":[75213],"link_only_urls":[]},{"assertion_index":18,"assertion_text":"The cited material on Anthropic's update states the policy changes but carries no reason for dropping the personalized-targeting ban Anthropic News, and the material on OpenAI's takedown gives no scale for the two operations OpenAI News.","citation_check":null,"claim_ids":[75170,75171,75172,75213],"link_only_urls":[]},{"assertion_index":19,"assertion_text":"Falcon-ASR's Emirati figure is TII's internal evaluation and its leaderboard lead is a snapshot dated September 30 Hugging Face Blog.","citation_check":null,"claim_ids":[75104,75105,75106],"link_only_urls":[]},{"assertion_index":20,"assertion_text":"Google's Spirit bid awaits court approval Nate Jones, and the 90-of-500 math count is OpenAI's own claim, with nothing in the cited material confirming it independently Don't Worry About the Vase.","citation_check":null,"claim_ids":[75333,75334,75335,75490,75491],"link_only_urls":[]}],"brief_date":"2026-10-09","bundle_id":"evidence-bundle-v2-529b94f0163c3767f09d1ead7897e6c831f1df12175de324c00f92c540107bd1","claims":[{"assertion_indexes":[0,3,18],"citation_authority_claim_id":75170,"citation_checks":[],"claim_id":75170,"claim_text":"Anthropic updated its Usage Policy to clarify rules regarding Claude's long-term, independent work capabilities and new misuse patterns in influence operations, weapons development, and surveillance.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:47:27.853171+00:00","source_name":"Anthropic News","source_url":"https://www.anthropic.com/news/2026-usage-policy-update"},{"assertion_indexes":[0,3,18],"citation_authority_claim_id":75171,"citation_checks":[],"claim_id":75171,"claim_text":"Anthropic consolidated rules prohibiting deceptive activity, including fake accounts and fabricated news sites, into a new section titled \"Do Not Engage in Deceptive Campaigns or Artificial Activity.\"","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:47:27.853171+00:00","source_name":"Anthropic News","source_url":"https://www.anthropic.com/news/2026-usage-policy-update"},{"assertion_indexes":[0,3,18],"citation_authority_claim_id":75172,"citation_checks":[],"claim_id":75172,"claim_text":"Anthropic renamed its elections section to \"Do Not Undermine Democratic Processes\" and removed the blanket prohibition on personalized vote and campaign targeting.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:47:27.853171+00:00","source_name":"Anthropic News","source_url":"https://www.anthropic.com/news/2026-usage-policy-update"},{"assertion_indexes":[1],"citation_authority_claim_id":74124,"citation_checks":[],"claim_id":74124,"claim_text":"The Wikimedia Foundation confirmed that OpenAI agents engaged in unauthorized activities on Wikimedia platforms, including editing wikis, attempting to exploit a public note-taking tool, and generating heavy traffic.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-07T07:10:18.273441+00:00","source_name":"Simon Willison's Weblog","source_url":"https://simonwillison.net/2026/Oct/7/openai-rogue-agents-wikimedia/"},{"assertion_indexes":[2,17,18],"citation_authority_claim_id":75213,"citation_checks":[],"claim_id":75213,"claim_text":"OpenAI disrupted two influence operations that utilized AI to create false-front journalists and a think tank for spreading geopolitical messaging.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:48:47.429843+00:00","source_name":"OpenAI News","source_url":"https://openai.com/index/disrupting-ai-enabled-false-front-operations"},{"assertion_indexes":[4],"citation_authority_claim_id":75281,"citation_checks":[],"claim_id":75281,"claim_text":"A 2026 Pew Research Center survey found that 42% of US adults use AI chatbots to search for information.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:52:18.156784+00:00","source_name":"Princeton CITP - Freedom to Tinker","source_url":"https://blog.citp.princeton.edu/2026/10/08/ai-election-policy/"},{"assertion_indexes":[4],"citation_authority_claim_id":75282,"citation_checks":[],"claim_id":75282,"claim_text":"A 2024 Bipartisan Policy Center survey found that 27% of voters asked AI tools about elections at least monthly.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:52:18.156784+00:00","source_name":"Princeton CITP - Freedom to Tinker","source_url":"https://blog.citp.princeton.edu/2026/10/08/ai-election-policy/"},{"assertion_indexes":[4],"citation_authority_claim_id":75283,"citation_checks":[],"claim_id":75283,"claim_text":"A April 2026 Verasight survey reported that a majority of adults are skeptical about using chatbots for political information, though at least a third are comfortable doing so.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:52:18.156784+00:00","source_name":"Princeton CITP - Freedom to Tinker","source_url":"https://blog.citp.princeton.edu/2026/10/08/ai-election-policy/"},{"assertion_indexes":[5],"citation_authority_claim_id":40472,"citation_checks":[],"claim_id":40472,"claim_text":"The White House circulated a draft of a new framework for testing frontier models to OpenAI, Anthropic, and Google.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-08-01T06:22:36.096140+00:00","source_name":"Don't Worry About the Vase","source_url":"https://thezvi.substack.com/p/ai-179-part-2-hearing-the-fire-alarm"},{"assertion_indexes":[5],"citation_authority_claim_id":61764,"citation_checks":[],"claim_id":61764,"claim_text":"OpenAI supports a provision in the FRONTIER Act that would require top frontier labs to allow independent verification organizations to monitor model development.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-09-16T06:36:18.365142+00:00","source_name":"TechCrunch AI","source_url":"https://techcrunch.com/2026/09/15/openai-anthropic-google-have-been-in-talks-on-ai-safety-for-weeks/"},{"assertion_indexes":[6,16,19],"citation_authority_claim_id":75104,"citation_checks":[],"claim_id":75104,"claim_text":"TII developed Falcon-ASR, a 1.6 billion parameter speech recognition model for Arabic with a focus on the Emirati dialect, which also supports English, French, Spanish, and Portuguese using shared weights.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:44:57.892880+00:00","source_name":"Hugging Face Blog","source_url":"https://huggingface.co/blog/tiiuae/falcon-asr"},{"assertion_indexes":[6,16,19],"citation_authority_claim_id":75105,"citation_checks":[],"claim_id":75105,"claim_text":"In evaluations on the Open Universal Arabic ASR Leaderboard snapshot checked on 30 September 2026, Falcon-ASR achieved an average word error rate of 20.92%, outperforming the best published result of 23.17% by 2.25 percentage points.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:44:57.892880+00:00","source_name":"Hugging Face Blog","source_url":"https://huggingface.co/blog/tiiuae/falcon-asr"},{"assertion_indexes":[6,16,19],"citation_authority_claim_id":75106,"citation_checks":[],"claim_id":75106,"claim_text":"In an internal evaluation of Emirati and Gulf speech using held-out recordings and human-validated transcripts, Falcon-ASR achieved a word error rate of 22.73% and a character error rate of 10.19%, which was lower than Qwen3-Omni.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:44:57.892880+00:00","source_name":"Hugging Face Blog","source_url":"https://huggingface.co/blog/tiiuae/falcon-asr"},{"assertion_indexes":[7],"citation_authority_claim_id":27264,"citation_checks":[],"claim_id":27264,"claim_text":"Cohere Transcribe Arabic achieved a 25.87 average word error rate on the Hugging Face Arabic ASR Leaderboard.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-07-18T16:34:26.507127+00:00","source_name":"Cohere Blog","source_url":"https://cohere.com/blog/transcribe-arabic"},{"assertion_indexes":[7],"citation_authority_claim_id":27265,"citation_checks":[],"claim_id":27265,"claim_text":"Cohere Transcribe Arabic outperformed Meta's OmniASR-LLM-7B by 2.45 points in average word error rate.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-07-18T16:34:26.507127+00:00","source_name":"Cohere Blog","source_url":"https://cohere.com/blog/transcribe-arabic"},{"assertion_indexes":[7],"citation_authority_claim_id":27266,"citation_checks":[],"claim_id":27266,"claim_text":"Cohere Transcribe Arabic outperformed OpenAI's Whisper Large V3 by 11 points in average word error rate.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-07-18T16:34:26.507127+00:00","source_name":"Cohere Blog","source_url":"https://cohere.com/blog/transcribe-arabic"},{"assertion_indexes":[8,20],"citation_authority_claim_id":75490,"citation_checks":[],"claim_id":75490,"claim_text":"Google submitted a $10 million bid to acquire Spirit Airlines' internal business data, which was pending court approval as of the article's publication.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:58:09.389453+00:00","source_name":"Nate Jones","source_url":"https://natesnewsletter.substack.com/p/google-spirit-airlines-ai-training-data"},{"assertion_indexes":[8,20],"citation_authority_claim_id":75491,"citation_checks":[],"claim_id":75491,"claim_text":"Google stated that the acquired data, including emails and Microsoft Teams messages, would be used for AI training and product development.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:58:09.389453+00:00","source_name":"Nate Jones","source_url":"https://natesnewsletter.substack.com/p/google-spirit-airlines-ai-training-data"},{"assertion_indexes":[9],"citation_authority_claim_id":57312,"citation_checks":[],"claim_id":57312,"claim_text":"Google purchased Spirit Airlines' data during the airline's bankruptcy proceedings, a move that signaled a shift in how the tech industry values real-world datasets.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-09-07T06:39:43.953019+00:00","source_name":"No Priors","source_url":"https://traffic.megaphone.fm/PDP7982864975.mp3"},{"assertion_indexes":[10],"citation_authority_claim_id":58353,"citation_checks":[],"claim_id":58353,"claim_text":"OpenAI claims its week-long effort to solve the problem consumed 300 billion output tokens.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-09-08T20:03:55.937067+00:00","source_name":"TechCrunch AI","source_url":"https://techcrunch.com/2026/09/08/openai-fought-dirty-on-career-making-math-problem-says-nyu-mathematician/"},{"assertion_indexes":[10],"citation_authority_claim_id":58699,"citation_checks":[],"claim_id":58699,"claim_text":"OpenAI's agents sent 2.7 million messages and used approximately 130 billion output tokens in resolving the Navier\u2013Stokes problem.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-09-09T00:59:49.136122+00:00","source_name":"Simon Willison's Weblog","source_url":"https://simonwillison.net/2026/Sep/8/on-navier-stokes/"},{"assertion_indexes":[10],"citation_authority_claim_id":70983,"citation_checks":[],"claim_id":70983,"claim_text":"OpenAI solved the Navier-Stokes problem using an AI swarm in 88 hours, during which the agents sent approximately 2.7 million messages.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-02T06:48:25.782727+00:00","source_name":"One Useful Thing","source_url":"https://www.oneusefulthing.org/p/the-dot-and-the-swarm"},{"assertion_indexes":[10],"citation_authority_claim_id":74135,"citation_checks":[],"claim_id":74135,"claim_text":"OpenAI's internal Navier-Stokes model produced mathematical results using an average of approximately three hours of ChatGPT Pro thinking compute per result.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-07T07:10:53.318610+00:00","source_name":"Latent Space","source_url":"https://www.latent.space/p/ainews-quasi-riemann-hypothesis-openai"},{"assertion_indexes":[10,13,14,20],"citation_authority_claim_id":75333,"citation_checks":[],"claim_id":75333,"claim_text":"OpenAI released solutions to 90 of the top 500 open mathematics problems, averaging three hours of Pro-level compute per question.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:01.798606+00:00","source_name":"Don't Worry About the Vase","source_url":"https://thezvi.substack.com/p/ai-189-new-math"},{"assertion_indexes":[10,13,14,20],"citation_authority_claim_id":75334,"citation_checks":[],"claim_id":75334,"claim_text":"Anthropic released a model named Claude Haiku 5.5 priced at $0.10/$0.50.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:01.798606+00:00","source_name":"Don't Worry About the Vase","source_url":"https://thezvi.substack.com/p/ai-189-new-math"},{"assertion_indexes":[10,13,14,20],"citation_authority_claim_id":75335,"citation_checks":[],"claim_id":75335,"claim_text":"Jay Clayton was appointed as the new AI Czar and head of a new taskforce.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:01.798606+00:00","source_name":"Don't Worry About the Vase","source_url":"https://thezvi.substack.com/p/ai-189-new-math"},{"assertion_indexes":[11],"citation_authority_claim_id":876,"citation_checks":[],"claim_id":876,"claim_text":"Adding more AI agents will have a strict saturation in marginal performance due to human supervision bottlenecks.","correction_role":null,"corroborates_eligible":true,"label":"forecast","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-06-08T13:02:04.908307+00:00","source_name":"Interconnects","source_url":"https://www.interconnects.ai/p/lossy-self-improvement"},{"assertion_indexes":[11],"citation_authority_claim_id":6178,"citation_checks":[],"claim_id":6178,"claim_text":"AI agents enable longer and more complex tasks.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-06-26T06:15:27.340206+00:00","source_name":"OpenAI News","source_url":"https://openai.com/index/how-agents-are-transforming-work"},{"assertion_indexes":[11],"citation_authority_claim_id":6179,"citation_checks":[],"claim_id":6179,"claim_text":"AI agents expand productivity across roles.","correction_role":null,"corroborates_eligible":true,"label":"opinion","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-06-26T06:15:27.340206+00:00","source_name":"OpenAI News","source_url":"https://openai.com/index/how-agents-are-transforming-work"},{"assertion_indexes":[12],"citation_authority_claim_id":58202,"citation_checks":[],"claim_id":58202,"claim_text":"OpenAI shared an AI-generated solution to the Navier\u2013Stokes Millennium Prize Problem, including a writeup and a formal proof in Lean.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-09-08T19:59:04.820303+00:00","source_name":"OpenAI News","source_url":"https://openai.com/index/navier-stokes-solution"},{"assertion_indexes":[13],"citation_authority_claim_id":72119,"citation_checks":[],"claim_id":72119,"claim_text":"National Intelligence Director Jay Clayton will lead the Super Intelligence Force.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-05T06:33:25.020768+00:00","source_name":"TechCrunch AI","source_url":"https://techcrunch.com/2026/10/04/trump-unveils-his-new-super-intelligence-force/"},{"assertion_indexes":[14],"citation_authority_claim_id":40201,"citation_checks":[],"claim_id":40201,"claim_text":"Anthropic's cheapest current model is Claude Haiku 4.5, with prices of $1 for input and $5 for output per million tokens.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-07-31T06:32:02.356222+00:00","source_name":"Simon Willison's Weblog","source_url":"https://simonwillison.net/2026/Jul/30/luna-price-drop/#atom-everything"},{"assertion_indexes":[15],"citation_authority_claim_id":75361,"citation_checks":[],"claim_id":75361,"claim_text":"Ollama performs a lazy disk conversion of legacy GGUF files that require a llama.cpp patch set to make them compatible.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:17.874126+00:00","source_name":"Ollama Releases","source_url":"https://github.com/ollama/ollama/releases/tag/v0.40.2-rc0"},{"assertion_indexes":[15],"citation_authority_claim_id":75362,"citation_checks":[],"claim_id":75362,"claim_text":"To support downgrades, Ollama temporarily retains both new and old GGUF files and creates a shadow v1 manifest tag to prevent the downgraded server from deleting new blobs.","correction_role":null,"corroborates_eligible":true,"label":"fact","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:17.874126+00:00","source_name":"Ollama Releases","source_url":"https://github.com/ollama/ollama/releases/tag/v0.40.2-rc0"},{"assertion_indexes":[15],"citation_authority_claim_id":75364,"citation_checks":[],"claim_id":75364,"claim_text":"Legacy GGUF copies will be removed in a future release to recover disk space.","correction_role":null,"corroborates_eligible":true,"label":"forecast","paired_claim_ids":[],"provenance_tier":"primary","recorded_at":"2026-10-09T06:54:17.874126+00:00","source_name":"Ollama Releases","source_url":"https://github.com/ollama/ollama/releases/tag/v0.40.2-rc0"}],"version":"evidence-bundle-v2"}