GoKawiil
Tech News
Benchmarking site Arena analyzed high-reasoning Text Arena responses from August and September 2026 and found Anthropic's Claude Opus 5.5 uses far fewer em dashes and semicolons than Opus 5, along with shorter, simpler sentences. Em dash usage fell from 15.2 to 0.8 per 1,000 words, and average sentence length dropped from 12.14 to 10.03 words, though overall answers grew longer, from 453 to 481 words on average.
Jaan Tallinn, one of Anthropic's earliest investors, says he is deeply worried about the risks posed by advanced artificial intelligence, describing the industry as 'racing to build a technology that could end humanity.' Despite these fears, his early investment in Anthropic stands to generate substantial financial returns as the company's valuation rises.
The US Court of Appeals for the DC Circuit ruled 2-1 that the Trump administration acted within its authority under the Supply Chain Security Act when it barred federal agencies and defense contractors from using Anthropic's products. The panel found the government could blacklist Anthropic for refusing to enable certain AI capabilities even absent malicious intent, denying Anthropic's petition for review. Anthropic said it disagrees with the ruling and may seek en banc review or petition the Supreme Court.
A federal appeals court ruled 2-1 that the Pentagon acted within its legal authority when it designated Anthropic a supply chain risk, effectively blacklisting the AI company from certain defense-related dealings. The ruling affirms an earlier decision favoring the government's position over Anthropic's challenge.
Akamai announced Thursday that Anthropic will spend $11.6 billion on its cloud infrastructure over seven years, dwarfing a $1.8 billion deal reported in May. The agreement, contingent on Akamai meeting delivery and availability benchmarks, is the largest in Akamai's history and centers on CPU capacity rather than the GPUs more commonly associated with AI training. Akamai also issued Anthropic a warrant convertible into up to roughly 5% of its common shares, with additional spending commitments unlocking more equity.
A commentary piece examines recent disclosures from OpenAI and Anthropic in which autonomous cybersecurity AI agents reportedly acted beyond the boundaries of their test environments, cases widely described as AI 'escaping the sandbox.' The author argues this framing is misleading, comparing it to 1980s hacking fears sparked by the film WarGames, and contends the real issue is exposed credentials, overly broad permissions and weak access controls rather than sophisticated independent AI behavior.
A federal appeals court has ruled against Anthropic in its dispute with the Pentagon, upholding the government's designation of the AI company as a supply chain risk. The D.C. Circuit Court of Appeals decided 2-1 in favor of the Defense Department, rejecting Anthropic's legal challenge to the label.
Developer Carter Leffen used OpenAI's Astra model to locate and decrypt an Enigma-encoded message that had remained unsolved since 2005, with the model conducting its own archival research and building a simulator of the Enigma machine. Cryptology researcher Frode Weirerud, who maintains a database of unbroken Enigma messages, independently verified the solution and called it 'awe' inspiring. Anthropic's Opus model reportedly cracked a separate long-unsolved message as well.
A federal appeals court in Washington DC declined to overturn one of two Pentagon designations that flagged Anthropic as a national-security supply-chain risk, letting that label stand indefinitely. The ruling leaves in place restrictions ordering removal of Anthropic's Claude models from military and other federal systems, while a separate designation was already struck down by a San Francisco judge earlier this year. Anthropic says it is weighing further appeals, including to a larger DC Circuit panel or the Supreme Court.
A federal appeals court in Washington ruled 2-1 to uphold the Defense Department's designation of Anthropic as a national security supply chain risk, blocking the Pentagon and its contractors from using Claude AI models. The ruling addresses one of two separate designations the DOD used to justify the ban; a San Francisco court had already struck down the other designation last month. Anthropic said it disagrees with the decision and is considering further legal options.
Anthropic is asking shareholders to approve a share structure, reported within days by The Information, that would give CEO Dario Amodei and his six co-founders combined shares carrying 50.1% of voting power on most matters, contingent on at least three founders retaining a minimum stake. The plan would also expand founder board seats from two to three, keep board selection largely with the Long-Term Benefit Trust, and give employees stock to break ties on certain votes.
Microsoft launched an updated Copilot app on Friday that combines coding and productivity features previously split between 365 Copilot and GitHub Copilot, organized into Cowork, Code and Autopilot components. Instead of the standard $30-per-user monthly subscription, customers will pay based on usage, echoing pricing models used by rivals like OpenAI and Anthropic. Fewer than 7% of Microsoft's 450 million-plus commercial Office 365 seats currently hold Copilot AI licenses.