Sam's News — anthropic — 2026-09-09¶

AI Safety¶

7.5 Anthropic safety researcher warns of 10%+ chance AI could cause human extinction by 2030s¶

A senior Anthropic safety researcher stated there is more than a 10% chance AI could cause human extinction by the 2030s, hours after researcher Jacob Coxon resigned over concerns about reckless AI development at Anthropic and OpenAI. Both researchers warned that companies are racing toward self-improving superintelligence without adequate safety measures.

10%+ estimated probability of AI-caused human extinction by end of decade
Anthropic researcher Jacob Coxon resigned citing safety concerns
Both Anthropic and OpenAI accused of prioritizing speed over safety protocols
Self-improving AI systems pose risk of spiraling beyond human control

Sources: The Verge AI Web Searched, tech.yahoo.com RSS, bbc.com RSS, BBC RSS, The Cryptonomist RSS, LADbible RSS

7.5 Anthropic Researcher Resigns Claiming OpenAI and Anthropic Are 'Gambling with Our Lives'¶

A researcher has quit Anthropic, alleging that both Anthropic and OpenAI are recklessly advancing AI despite existential risks.

Sources: Business Insider RSS, Yahoo Tech RSS, TheGrio RSS, BBC RSS, foxbusiness.com RSS, cbsnews.com RSS

AI¶

7 Anthropic Unveils Claude Commerce Agents; Partners With Visa and Mastercard¶

Anthropic has launched Claude Commerce Agents in partnership with Visa and Mastercard for transaction processing.

Sources: crowdfundinsider.com RSS

6.5 Anthropic Releases New Model to Test AI's Economic Impact on the U.S.¶

Anthropic has developed a new model to assess and test how artificial intelligence could affect the U.S. economy.

Sources: npr.org RSS

6.5 Salesforce's $300M Annual Claude Spending Triggers Cost Optimization; Company Cuts Token Costs¶

Salesforce's substantial annual Claude spending of $300 million has forced the company to implement cost-reduction measures and negotiate lower token pricing.

Sources: 36 Kr RSS

Security¶

7 Claude AI accounts compromised; stolen API tokens reported¶

Users of Anthropic's Claude have reported compromised account sessions resulting in stolen AI API tokens.

Sources: Asaase Radio RSS

6.5 Anthropic Faces Questions Over Claude Security Breaches and AI Safety Implications¶

Security breaches affecting Anthropic's Claude AI model raise questions about the company's approach to AI safety.

Sources: Tempo.co English RSS

Tech¶

6.5 Anthropic Did Not Submit Strongest AI Model for UK Safety Testing¶

Anthropic did not submit its most powerful AI model for evaluation in the UK regulatory testing process.

Sources: The Times RSS

Funding¶

6.5 Anthropic Races Toward $2 Trillion IPO With $1 Trillion Revenue Target by 2030¶

Investors bullish on Anthropic project the company could reach a $2 trillion valuation with $1 trillion in annual revenue by 2030.

Sources: finance.biggo.com RSS

AI Business¶

6.5 Salesforce adopts model-routing strategy to manage Claude API costs¶

Salesforce is using model-routing techniques to address its substantial Claude API spending and optimize AI infrastructure costs.

Sources: finance.biggo.com RSS, 36Kr RSS