Skip to content

Sam's News — anthropic — 2026-09-28

AI

7.5 Anthropic Releases Claude Sonnet 5.5 Model

Anthropic released Claude Sonnet 5.5 on September 28, delivering 30% faster inference and 30% lower cost per task despite identical token pricing. The model scores 70.6% on Terminal-Bench 4.0 agentic coding versus Sonnet 5's 10.3%, and is the first Sonnet to beat Pokémon Red using only screenshots.

  • Released September 28; 30%+ faster than Sonnet 5, 30% cheaper per task
  • Token pricing unchanged: $2M input, $10M output, $0.20M cache read
  • Scores 70.6% Terminal-Bench 4.0 vs Sonnet 5's 10.3%
  • First Sonnet to beat Pokémon Red; includes cybersecurity safeguards comparable to Opus 5.5

Sources: Anthropic AI Web Searched, Hacker News (front page) RSS, The GitHub Blog RSS

7 Anthropic's Claude AI solves advanced particle physics problem

Anthropic announced that Claude AI successfully solved a nine-loop particle physics challenge.

Sources: Dataconomy RSS Update to: Claude completes nine-loop scattering amplitude calculation, breaking three-year theoretical physics record

6.5 Claude Opus 5.5 beats GPT-5.6 Sol at one-third the cost

Anthropic's Claude Opus 5.5 demonstrated superior performance compared to OpenAI's GPT-5.6 Sol while being significantly cheaper to operate.

Sources: shattered.io RSS Update to: Anthropic releases Claude Opus 5.5 with 40% lower operating costs

6 Scrutiny Over Anthropic AI's Claimed Independent Scientific Discovery

The New York Times examines whether Anthropic's AI actually made a scientific discovery autonomously.

Sources: The New York Times RSS

6 The president of the United States meets Anthropic CEO following Pentagon AI safety disagreement

The president of the United States met with Anthropic's CEO after a dispute with the Pentagon over AI safety policy.

Sources: KUTV RSS, The Washington Post RSS

Policy

7 The president of the United States dines with Anthropic CEO Dario Amodei after AI policy tensions

The president of the United States dined privately with Anthropic CEO Dario Amodei at the White House on September 28, marking their first reported one-on-one meeting after months of conflict over military use of Claude AI models. The dinner occurred two days after a federal appeals court upheld Pentagon exclusion of Anthropic as a supply-chain risk, amid tensions over AI oversight and capability growth.

  • Private White House dinner September 28 between president and Amodei
  • First one-on-one meeting after months of AI policy conflict
  • Tension: Anthropic advocates greater AI oversight; president resists restrictions
  • Federal appeals court upheld Pentagon decision excluding Anthropic as supply-chain risk two days prior

Sources: New York Post AI Web Searched

6.5 President of the United States Plans Dinner with Anthropic CEO Dario Amodei

The president of the United States announced plans to dine with Anthropic CEO Dario Amodei amid ongoing debates over AI safety and policy.

Sources: Sowetan RSS, ABC News - Breaking News, Latest News and Videos RSS, The Times of India RSS

6.5 Anthropic loses Pentagon contract appeal over Claude's safety safeguards

Anthropic's appeal to reverse a Pentagon blacklisting over Claude's safety features has been denied.

Sources: The Rundown AI RSS Update to: US court backs Pentagon's decision to exclude Anthropic from defense contracts

Regulation

6.5 US appeals court upholds Pentagon's national security designation for Anthropic

A federal appeals court affirmed the Pentagon's classification of Anthropic as a critical national security supply chain risk.

Sources: Digital Watch Observatory RSS Update to: Court backs Pentagon decision to label Anthropic a supply risk

Business

5.5 Anthropic Expands AI Hiring in India Following Claude Opus 5.5 Launch

Anthropic announced significant hiring expansion in India following the launch of Claude Opus 5.5.

Sources: dqindia.com RSS