Loading your language..
Anthropic announced that its own model Claude is assisting in the development of a next-generation version.

Anthropic announced that its own model Claude is assisting in the development of a next-generation version.

C2🇯🇵 日本語 → 🇺🇞 English

September 18th, 2026

Anthropic announced that its own model Claude is assisting in the development of a next-generation version.

C2
Please note: This article has been simplified for language learning purposes. Some context and nuance from the original text may have been modified or removed.

🇺🇞 English

In an announcement on Thursday, the AI research institute stated that Anthropic's Claude has been instrumental in the company's development of next-generation models endowed with more advanced intelligence.

AnthropicのClaudeは、同瀟が次䞖代の、より高床な知胜を備えたモデルを開発する䞊で助力ずなっおいるず、このAI研究所は朚曜日の発衚で蚀明した。

Claude is spearheading 26% of Anthropic's model research and development, which, the company stated, means that it can carry most tasks through to completion "end-to-end from a high-level prompt," albeit under human supervision.

ClaudeはAnthropicのモデル研究開発の26%を䞻導しおおり、これは人間の監督䞋にありながら、「高レベルなプロンプトから゚ンドツヌ゚ンドで」ほずんどのタスクを完遂し埗るこずを意味するず同瀟は述べた。

This model does not yet operate in a fully autonomous manner.

このモデルは未だ完党に自埋的に皌働しおいるわけではない。

Nevertheless, approximately 90% of the company's research and development is carried out through "collaboration" with Claude, which Anthropic stated means that the model can shoulder "the majority of the work under meticulous human direction."

にもかかわらず、同瀟の研究開発のおよそ90%はClaudeずの「協働」によっお遂行されおおり、このこずは、同モデルが「人間の綿密な指瀺の䞋で䜜業の倧郚分」を担い埗るこずを意味するず、Anthropicは述べた。

This announcement was made under the leadership of Dario Amodei, the CEO of Anthropic, at a time when some of the leading figures in the AI world were calling for a slowdown in technological development on the grounds of safety concerns.

本発衚は、AnthropicのCEOたるダリオ・アモデむ氏が䞻導する圢で、AI界の䞀郚の䞻芁人物らが安党䞊の懞念を理由に技術開発の枛速を呌びかける䞭で行われたものである。

In deliberating the pace of AI development, the company stated in a blog post that "every possible measure should be taken to minimize the gap that arises between what frontier labs know and what the general public knows."

AIの開発ペヌスを怜蚎するにあたり、同瀟はブログ投皿においお、「フロンティアラボが把握しおいる事柄ず䞀般公䌗が把握しおいる事柄ずの間に生じる隔たりを最小限に抑止すべく、あらゆる可胜な措眮を講じるべきである」ず述べた。

"This means measuring AI progress more precisely, reporting it publicly, and affording society the opportunity to decide how such information should be used."

「これはすなわち、AIの発展をより粟緻に枬定し、それを公に報告し、かかる情報をいかに掻甚するかに぀いお瀟䌚が決定を䞋す機䌚を付䞎するこずを意味する。」

The company stated that models accelerating their own development carries the potential to "make it more difficult for humans to understand and control these systems."

同瀟は、モデルが自らの開発を加速させるこずが「人間がこれらのシステムを理解し、制埡するこずをより困難にする」可胜性を孕んでいるず述べた。

It argued that sharing these metrics could enable a more precise grasp of how close leading AI labs are coming to recursive self-improvement—that is, the capacity for models to autonomously construct their successors.

これらの指暙を共有するこずにより、䞻芁なAIラボが再垰的自己改善——すなわちモデルが自埋的に埌継モデルを構築する胜力——にどれほど接近し぀぀あるかを、より粟緻に把握できるようになる可胜性があるず䞻匵した。

Anthropic further urged other AI developers to share similar metrics on a regular basis, and encouraged the adoption of a public methodology so as to enable comparisons over time and, potentially, across labs.

Anthropicはさらに、他のAI開発者に察しおも同様の指暙を定期的に共有するよう促すずずもに、経時的な、ひいおは朜圚的にはラボ間での比范を可胜ならしめるべく、公的な方法論の採甚を奚励した。

While Anthropic's disclosures shed little light on how close the company believes it is to achieving recursive self-improvement, the fact that Claude's contribution to research and development is growing at an accelerating rate is worthy of note.

Anthropicの開瀺からは、同瀟が再垰的自己改善の達成にどれほど接近しおいるず芋なしおいるかは刀然ずしなかったが、Claudeが研究開発に寄䞎する割合が加速床的に増倧しおいる事実は泚目に倀する。

The proportion of projects that Claude "leads" — that is, those in which it remains under human supervision yet carries out the bulk of the work — stood at zero in February.

Claudeが「䞻導」する、すなわち人間の監督䞋に眮かれ぀぀もその䜜業の倧郚分を遂行する案件の割合は、2月には皆無であった。

Six months later, it had come to lead a quarter of research and development work, reaching that level in August.

それから6か月埌には研究開発䜜業の4分の1を䞻導するに至り、8月にその氎準に達したのである。

The company further disclosed details of the agent oversight measures it has implemented, noting that as of August, approximately 30,000 agents were engaged in research and engineering work.

同瀟はさらに、導入枈みの゚ヌゞェント監督措眮の詳现を開瀺し、8月時点で玄3䞇の゚ヌゞェントが研究および゚ンゞニアリング業務に埓事しおいるず指摘した。

According to the company, these oversight measures are of paramount importance in verifying how frequently misconduct by agents is detected by the monitoring system.

同瀟によれば、圓該監督措眮は、゚ヌゞェントによる䞍正行為が監芖システムによっおいかなる頻床で怜出されるかを怜蚌する䞊で極めお重芁であるずいう。

Anthropic recently committed to stationing external third-party evaluators on-site within the company in order to oversee its safety efforts.

Anthropicは最近、安党ぞの取り組みを監芖すべく、瀟内に垞駐する倖郚の第䞉者評䟡者を蚭眮するこずを玄束した。

Last week, the resignation of an Anthropic researcher who left behind a grave warning about the threat this technology poses to humanity set in motion much of the recent discourse surrounding AI safety.

先週、Anthropicの研究者が、この技術が人類にもたらす脅嚁に関し深刻な譊告を遺しお蟞任したこずが、AI安党性をめぐる昚今の察話の倚くを端緒づけた。

Amodei, Sam Altman in his capacity as OpenAI's CEO, Elon Musk, and other technology leaders have since come to endorse the notion of slowing development, whereas other technology leaders, along with President Donald Trump, oppose it.

アモデむ氏、OpenAIのCEOたるサム・アルトマン氏、むヌロン・マスク氏、その他の技術指導者たちはその埌、開発の枛速ずいう構想を支持するに至っおいるが、他の技術指導者たちおよびドナルド・トランプ倧統領はこれに反察しおいる。

September 18th, 2026

Trending Articles

Warren Buffett to Step Down as Berkshire Hathaway Chairman: The Next Stage of the Succession Plan

Warren Buffett to Step Down as Berkshire Hathaway Chairman: The Next Stage of the Succession Plan

りォヌレン・バフェット、バヌクシャヌ・ハサりェむ䌚長職を退任ぞ埌継者蚈画の次段階

C2Sep 18
Trump says CNN, MS NOW, and Politico are barred from the White House, targeting further news organizations

Trump says CNN, MS NOW, and Politico are barred from the White House, targeting further news organizations

トランプ倧統領、CNN、MS NOW、ポリティコのホワむトハりス立ち入り犁止を発蚀、さらなる報道機関を暙的に

C2Sep 18
Bank of Japan raises policy rate to 1.25%, highest in 31 years

Bank of Japan raises policy rate to 1.25%, highest in 31 years

日銀、政策金利を1.25%ぞ匕き䞊げ、31幎ぶりの高氎準に

C2Sep 18
The Reasons and Implications of the FRB's Current Decision to Raise Interest Rates

The Reasons and Implications of the FRB's Current Decision to Raise Interest Rates

FRBが今、利䞊げに螏み切る理由ずその意味

C2Sep 17
US stocks surge most in six weeks amid falling oil prices and declining bond yields

US stocks surge most in six weeks amid falling oil prices and declining bond yields

米囜株、原油安ず債刞利回り䜎䞋を受け6週間ぶり最倧の急隰

C2Sep 17
Fed Raises Key Interest Rate for First Time in Three Years, Defying Trump's Demand for Cuts

Fed Raises Key Interest Rate for First Time in Three Years, Defying Trump's Demand for Cuts

FRB、3幎ぶりに䞻芁金利を匕き䞊げ——利䞋げを求めるトランプ倧統領の芁求に反する決断

C2Sep 17
Unemployment insurance claims fall to 196,000—lowest level since mid-July, layoffs remain subdued

Unemployment insurance claims fall to 196,000—lowest level since mid-July, layoffs remain subdued

倱業保険申請件数、19侇6000件ぞ枛少—7月䞭旬以来の最䜎氎準、解雇なお䜎調

C2Sep 17
President Trump dismisses AI risks as a "hoax," claims a "pathological conspiracy" surrounding AI and data centers

President Trump dismisses AI risks as a "hoax," claims a "pathological conspiracy" surrounding AI and data centers

トランプ倧統領、AIのリスクを「デマ」ず䞀蹎し、AIずデヌタセンタヌを巡る「病的な陰謀」の存圚を䞻匵

C2Sep 14
Supreme Court to Decide Whether to Hear Maine Lobstermen's Challenge to GPS Boat-Tracking Requirement

Supreme Court to Decide Whether to Hear Maine Lobstermen's Challenge to GPS Boat-Tracking Requirement

Supreme Court to Decide Whether to Hear Maine Lobstermen's Challenge to GPS Boat-Tracking Requirement

C2Sep 14