2026-11-05 notis
OTHER NOTIS SITES AIYOU ARE HERE AI models, prices and agents. ainotis.com Hiking Trail openings, hut bookings, and route conditions. hiking.notis.news Photography Cameras, lenses and firmware. photo.notis.news More areas when we can cover them every day.
CATEGORY Join
Show Read as Find Sections Models and releasesAgents and appsResearchMoney and computeLaws and rulingsSafety and securityAI at work More ArchiveModelsAboutCorrectionsThe numbersRSS
My notis
TODAY'S NEWS
November 2026 3 STORIES
MTWTFSS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Days with a dot have news. See the archive.

The whole week ›
16:37
0/3
WEATHER THEME CARD SPACING

Today's news

today

My page

Kept in this browser only.

SAVED FOLLOWING COLLECTIONS READING SINCE
SAVED STORIES
Nothing saved yet. Tap the bookmark on a card to keep it here. read today
CASES YOU FOLLOW all cases

No case followed yet.

COLLECTIONS
Price changes Stories where a published price went up or down.
Models being retired Models being withdrawn, and gaps before a replacement is ready.
THE MORNING EMAIL The day's top stories. Sign up
Safety 5 NOV

Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold

AnthropicIPOResponsible Scaling Policy
Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023.Kimberly White/Getty Images for TechCrunch / Dario Amodei at TechCrunch Disrupt 2023 / CC BY 2.0, cropped

Amendment No. 1 to its S-1, filed 2 November, rests the statement on estimates for future models; Anthropic says it has designated no current model as exceeding the threshold.

SOURCEU.S. SEC / Anthropic, PBC · 4 sources
CHECKED5 Nov, 05:21
TOPICSFrontier risk disclosures
Safety read
Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023. Safety / 5 NOV

Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold
  • Anthropic's S-1 amendment, filed 2 November 2026, says its estimates for future models indicate it may imminently meet its RSP definition of "dramatically accelerated progress attributable to AI."
  • The S-1 Anthropic filed on 26 October contains neither "imminently" nor "dramatically accelerated progress", so the wording is new in the amendment.
  • Anthropic says it has not designated any current model as exceeding the threshold, continues to assess the data, and will release further reports as its RSP sets out.

Anthropic says in an amended registration statement that it "may imminently meet" the definition of "dramatically accelerated progress attributable to AI" in its Responsible Scaling Policy. The wording is new in Amendment No. 1 to its Form S-1, filed with the US Securities and Exchange Commission on 2 November 2026; it does not appear in the S-1 filed on 26 October.

The filing ties the statement to internal estimates and to expectations about future models rather than current ones. It says Anthropic has not designated any of its current models as exceeding the threshold and that it continues to assess the underlying data.

Anthropic also says it is evaluating what meeting the AI research and development capability threshold would mean, and that it will release further reports as its Responsible Scaling Policy sets out.

The policy's version 3.4, effective 8 July 2026, says the automated R&D threshold is met if Anthropic determines that its models could fully substitute for its entire set of research scientists and research engineers at competitive costs, or that there is "dramatic acceleration" of the pace of AI progress for reasons that likely relate to automating AI research.

The same amendment fills in financial figures the first filing left blank, including aggregate committed spend of about $895.4 billion, substantially all for compute and infrastructure.

What it means for you Our view

This is a statement about expectations for future models, not a finding about any model Anthropic sells today: the filing says no current model is designated past the threshold. What it adds is that a company preparing to sell shares tells investors the trigger in its own scaling policy may be close, in wording the first filing did not contain. The filing also says Anthropic is still assessing the underlying data, so the position may change. If you depend on Anthropic's models or on its published safety commitments, watch for the further reports the filing says Anthropic will release in the future under its Responsible Scaling Policy.

TAGSAnthropicIPOResponsible Scaling Policy
U.S. SEC / Anthropic, PBC · 4 sources tap to close
Models 5 NOV

Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first

Hugging FaceKimiKimi licenceMoonshot AIOpen weights
Illustration: A tall steel cabinet with its door open, a dial counter on its side and a padlock plate by the frame. The green door stands for released weights.Illustration made with AI for ai notis

The weights are ungated on Hugging Face; a company running a model-as-a-service business, if it and its affiliates take in more than $20 million over any 12 months, needs a Moonshot agreement.

SOURCEMoonshot AI (Hugging Face) · 3 sources
CHECKED5 Nov, 05:21
TOPICSOpen weights at frontier scale, Permissive licences on the largest open weights
Models read
Illustration: A tall steel cabinet with its door open, a dial counter on its side and a padlock plate by the frame. The green door stands for released weights. Models / 5 NOV

Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first
  • Moonshot has put Kimi K3.1 Pro on Hugging Face, ungated, in 96 weight files: 2.8T total parameters, 104B active, and a context length of 1048576 tokens.
  • A company running a model-as-a-service business whose aggregate revenue, with its affiliates, exceeds 20 million US dollars over any 12 months must sign a separate Moonshot AI agreement first.
  • Products with over 100 million monthly active users, or 20 million US dollars in monthly revenue, must display "Kimi K3.1-Pro"; internal use and Moonshot's own products or partners are exempt.

Moonshot AI has put the weights of Kimi K3.1 Pro on Hugging Face, a mixture-of-experts model with 2.8 trillion parameters in total, 104 billion active per token, and a context length of 1,048,576 tokens. The repository, which Hugging Face's API lists as created on 2 November 2026, is not gated and holds 96 weight files.

Moonshot's model card calls it an open-weight, natively multimodal agentic model that takes text and images, and says it was post-trained from Kimi K3. Moonshot's tech blog had said the full weights would be released by 4 November.

The licence is Moonshot's own, not a standard open-source one. It lets anyone use, copy, modify and sell the software, but if a licensee or its affiliates run a model-as-a-service business and their aggregate revenue exceeds 20 million US dollars over any consecutive 12 months, the licensee must sign a separate agreement with Moonshot before using the software for any commercial purpose.

Products with more than 100 million monthly active users, or more than 20 million dollars in monthly revenue, must display "Kimi K3.1-Pro" on the interface. Internal use, and use through Moonshot's own products or certified inference partners, are exempt from both conditions.

Moonshot also lists the weights as MXFP4 weights with MXFP8 activations, trained with quantization-aware training.

What it means for you Our view

The licence grants broad rights to use, copy, modify and sell the weights, and its conditions are aimed at businesses that offer model access to others. If you run Kimi K3.1 Pro only for your own staff, the internal-use exemption covers use that does not make the software, its outputs or its capabilities available to third parties, so a customer-facing feature needs a closer look. Use through Moonshot's official products or certified inference partners is exempt too. Before you build on the weights, have your legal team read sections 2 to 4 of the licence against the product you plan to ship.

TAGSHugging FaceKimiKimi licenceMoonshot AIOpen weights
Moonshot AI (Hugging Face) · 3 sources tap to close
Compute 5 NOV

AMD values its pending Taalas inference-chip purchase at about $2.9 billion

AMDAcquisitionsInference chipsTaalas
Lisa Su, AMD's chief executive, photographed at South by Southwest in Austin, Texas, in March 2024.Fuzheado / SXSW-2024-alih-OB7A0861-Lisa Su / CC BY 4.0, cropped

Taalas develops specialised AI inference silicon; up to about $1.4 billion of the price is contingent, and AMD's 10-Q expects the deal to close by the end of 2026.

SOURCEU.S. SEC / AMD · 2 sources
CHECKED5 Nov, 05:21
TOPICSInference-only accelerators
Compute read
Lisa Su, AMD's chief executive, photographed at South by Southwest in Austin, Texas, in March 2024. Compute / 5 NOV

AMD values its pending Taalas inference-chip purchase at about $2.9 billion
  • AMD's 10-Q, for the quarter ended 26 September 2026, values the cash-and-stock acquisition of Taalas at about $2.9 billion, including contingent consideration of up to about $1.4 billion.
  • AMD announced a definitive agreement on 6 August 2026 in a press release that stated no price; the release calls Taalas a pioneer in specialized AI inference silicon.
  • The 10-Q says the acquisitions are expected to close by the end of 2026, subject to regulatory approvals and customary closing conditions.

AMD's quarterly filing puts a price on its planned purchase of the inference-chip maker Taalas: about $2.9 billion in cash and stock, including contingent consideration of up to about $1.4 billion. The figure is in Note 5 of AMD's Form 10-Q for the quarter ended 26 September 2026, filed on 4 November.

The deal is not closed. The 10-Q describes it as a pending acquisition that is expected to close by the end of 2026, subject to regulatory approvals and customary closing conditions.

AMD first announced a definitive agreement on 6 August 2026 without stating a price. The press release describes Taalas as a Toronto-based developer of specialised AI inference silicon, founded in 2023, and says AMD plans to integrate the technology into its accelerator roadmap alongside AMD Instinct GPUs.

The 10-Q also reports Data Center net revenue of $8.5 billion for the three months ended 26 September 2026, up 97% from $4.3 billion a year earlier. It attributes the rise primarily to demand for AMD EPYC processors and AMD Instinct GPUs.

AMD's future purchase commitments as of 26 September 2026 total $48,761 million, of which $11,218 million falls in the remainder of 2026 and $22,206 million in 2027.

What it means for you Our view

The price is a filing disclosure about a pending deal, so nothing changes yet in what AMD sells. AMD says it plans to integrate Taalas's technology into its accelerator roadmap and to develop system-level solutions with AMD Instinct GPUs, which is a plan and not a product. If you are weighing inference hardware, treat anything Taalas-based from AMD as something to watch rather than to buy, and note the year-end closing expectation as the date to check for news on regulatory approval.

TAGSAMDAcquisitionsInference chipsTaalas
U.S. SEC / AMD · 2 sources tap to close
NOTHING IN THIS FILTER No story today matches. open the archive →
That's today's news: three stories. ExplainedHow Anthropic's RSP decides a capability threshold is met, and what follows

The morning email

On the mornings we publish: the three top stories and up to four short ones. Free.

We email you a link to confirm. An issue may include one sponsor, always labelled Sponsored · Advertisement. Our emails count opens and clicks, not who made them. Unsubscribe in one click. What we keep

1 of 3
Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023. Kimberly White/Getty Images for TechCrunch / Dario Amodei at TechCrunch Disrupt 2023 / CC BY 2.0, cropped

Safety · 5 NOV

Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold

Amendment No. 1 to its S-1, filed 2 November, rests the statement on estimates for future models; Anthropic says it has designated no current model as exceeding the threshold.

  • Anthropic's S-1 amendment, filed 2 November 2026, says its estimates for future models indicate it may imminently meet its RSP definition of "dramatically accelerated progress attributable to AI."
  • The S-1 Anthropic filed on 26 October contains neither "imminently" nor "dramatically accelerated progress", so the wording is new in the amendment.
  • Anthropic says it has not designated any current model as exceeding the threshold, continues to assess the data, and will release further reports as its RSP sets out.
Checked 5 Nov, 05:21 · U.S. SEC / Anthropic, PBC · 4 sources
Read the story
Illustration: A tall steel cabinet with its door open, a dial counter on its side and a padlock plate by the frame. The green door stands for released weights. Illustration made with AI for ai notis

Models · 5 NOV

Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first

The weights are ungated on Hugging Face; a company running a model-as-a-service business, if it and its affiliates take in more than $20 million over any 12 months, needs a Moonshot agreement.

  • Moonshot has put Kimi K3.1 Pro on Hugging Face, ungated, in 96 weight files: 2.8T total parameters, 104B active, and a context length of 1048576 tokens.
  • A company running a model-as-a-service business whose aggregate revenue, with its affiliates, exceeds 20 million US dollars over any 12 months must sign a separate Moonshot AI agreement first.
  • Products with over 100 million monthly active users, or 20 million US dollars in monthly revenue, must display "Kimi K3.1-Pro"; internal use and Moonshot's own products or partners are exempt.
Checked 5 Nov, 05:21 · Moonshot AI (Hugging Face) · 3 sources
Read the story
Lisa Su, AMD's chief executive, photographed at South by Southwest in Austin, Texas, in March 2024. Fuzheado / SXSW-2024-alih-OB7A0861-Lisa Su / CC BY 4.0, cropped

Compute · 5 NOV

AMD values its pending Taalas inference-chip purchase at about $2.9 billion

Taalas develops specialised AI inference silicon; up to about $1.4 billion of the price is contingent, and AMD's 10-Q expects the deal to close by the end of 2026.

  • AMD's 10-Q, for the quarter ended 26 September 2026, values the cash-and-stock acquisition of Taalas at about $2.9 billion, including contingent consideration of up to about $1.4 billion.
  • AMD announced a definitive agreement on 6 August 2026 in a press release that stated no price; the release calls Taalas a pioneer in specialized AI inference silicon.
  • The 10-Q says the acquisitions are expected to close by the end of 2026, subject to regulatory approvals and customary closing conditions.
Checked 5 Nov, 05:21 · U.S. SEC / AMD · 2 sources
Read the story

You're all caught up

That's today's news: three stories.

ExplainedHow Anthropic's RSP decides a capability threshold is met, and what follows
Start again See the whole week

You can join ai notis soon, to follow cases and keep stories on every device.

Follow us on Instagram and Threads

The morning email

On the mornings we publish: the three top stories and up to four short ones. Free.

We email you a link to confirm. An issue may include one sponsor, always labelled Sponsored · Advertisement. Our emails count opens and clicks, not who made them. Unsubscribe in one click. What we keep

No story today matches. Open the archive
Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023. Kimberly White/Getty Images for TechCrunch / Dario Amodei at TechCrunch Disrupt 2023 / CC BY 2.0, cropped Safety · 5 NOV Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold U.S. SEC / Anthropic, PBC · 4 sources · 5 Nov, 05:21 Illustration: A tall steel cabinet with its door open, a dial counter on its side and a padlock plate by the frame. The green door stands for released weights. Illustration made with AI for ai notis Models · 5 NOV Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first Moonshot AI (Hugging Face) · 3 sources · 5 Nov, 05:21 Lisa Su, AMD's chief executive, photographed at South by Southwest in Austin, Texas, in March 2024. Fuzheado / SXSW-2024-alih-OB7A0861-Lisa Su / CC BY 4.0, cropped Compute · 5 NOV AMD values its pending Taalas inference-chip purchase at about $2.9 billion U.S. SEC / AMD · 2 sources · 5 Nov, 05:21

That's today's news: 3 stories.Archive

Today's stories 3 STORIES

Today's stories

Thu 5 Nov · 3 STORIES
  1. Safety · 5 NOV

    Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold

    Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023.Kimberly White/Getty Images for TechCrunch / Dario Amodei at TechCrunch Disrupt 2023 / CC BY 2.0, cropped

    Amendment No. 1 to its S-1, filed 2 November, rests the statement on estimates for future models; Anthropic says it has designated no current model as exceeding the threshold.

    Checked 5 Nov, 05:21 · U.S. SEC / Anthropic, PBC · 4 sourcesRead

  2. Models · 5 NOV

    Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first

    Illustration: A tall steel cabinet with its door open, a dial counter on its side and a padlock plate by the frame. The green door stands for released weights.Illustration made with AI for ai notis

    The weights are ungated on Hugging Face; a company running a model-as-a-service business, if it and its affiliates take in more than $20 million over any 12 months, needs a Moonshot agreement.

    Checked 5 Nov, 05:21 · Moonshot AI (Hugging Face) · 3 sourcesRead

  3. Compute · 5 NOV

    AMD values its pending Taalas inference-chip purchase at about $2.9 billion

    Lisa Su, AMD's chief executive, photographed at South by Southwest in Austin, Texas, in March 2024.Fuzheado / SXSW-2024-alih-OB7A0861-Lisa Su / CC BY 4.0, cropped

    Taalas develops specialised AI inference silicon; up to about $1.4 billion of the price is contingent, and AMD's 10-Q expects the deal to close by the end of 2026.

    Checked 5 Nov, 05:21 · U.S. SEC / AMD · 2 sourcesRead

Safety 5 NOV

Anthropic estimates its future models may imminently meet its RSP AI-acceleration threshold

Amendment No. 1 to its S-1, filed 2 November, rests the statement on estimates for future models; Anthropic says it has designated no current model as exceeding the threshold.

U.S. SEC / Anthropic, PBC · 4 sources·checked 5 Nov, 05:21
AnthropicIPOResponsible Scaling Policy
Dario Amodei, Anthropic's co-founder and chief executive, speaking on stage at TechCrunch Disrupt in 2023.Kimberly White/Getty Images for TechCrunch / Dario Amodei at TechCrunch Disrupt 2023 / CC BY 2.0, cropped
Models

Moonshot opens Kimi K3.1 Pro weights; model-service firms over $20M must sign first

The weights are ungated on Hugging Face; a company running a model-as-a-service business, if it and its affiliates take in more than $20 million over any 12 months, needs a Moonshot agreement.

5 NOV · Moonshot AI (Hugging Face) · 3 sources
Compute

AMD values its pending Taalas inference-chip purchase at about $2.9 billion

Taalas develops specialised AI inference silicon; up to about $1.4 billion of the price is contingent, and AMD's 10-Q expects the deal to close by the end of 2026.

5 NOV · U.S. SEC / AMD · 2 sources

Explained

One idea behind the news
Safety How Anthropic's RSP decides a capability threshold is met, and what follows Anthropic's Responsible Scaling Policy v3.4 counts its automated AI R&D threshold as met in two ways, and ties reports, external review and a conditional delay to it.
In brief 4 more today
0 of 3 read
Drag a card to start Drag left or right · tap to read more Hover to lift a card · tap to bring it forward Tap again to put it back Tap a card to open it Tap again to close it Tap a headline to read it as a card