Updated Jul 24
Nov 11, 2026
Anthropic's Fable 5.2 Completes 41% of Real Freelance Jobs, Up From 2.5% a Year Ago
Anthropic's Fable 5.2 scored 41.3% on the Remote Labor Index, completing nearly half of 240 real Upwork projects at professional quality. That's a 16x leap from the 2.5% top score a year ago and more than double the next-best system. The benchmark, run by Scale Labs and CAIS, uses human experts to grade deliverables — and nearly 50 of Fable 5.2's successes were tasks no AI had ever solved before.
anthropicclaudefable-5-2
Nov 11, 2026
New York's AI Safety Law Opens Registration: OpenAI and Anthropic Must Sign Up by January
New York opened its RAISE Act registration portal on Tuesday, requiring large frontier AI developers — including OpenAI and Anthropic — to register by January 1, 2027. Companies with over $500 million in annual revenue must publish safety frameworks, report critical incidents within 72 hours, and issue transparency reports. Fines reach $3 million for repeat violations.
new-yorkraise-actai-safety
Nov 11, 2026
Anthropic IPO Targets $1.4 Trillion Valuation, $64 Billion Raise
Anthropic set a $70 to $78 share price range for its IPO, valuing the Claude maker at roughly $1.4 trillion on a fully diluted basis. The offering could raise up to $64 billion, with Amazon and Google holding the biggest paper gains. Trading is expected to start November 19 under the ticker ANT.
anthropicanthropic-ipoclaude
Related News
Nov 11, 2026
AgentCore Web Answer vs Web Search: who runs the model, what you get back, and what the managed tool decides for you
AWS announced general availability of Web Answer on Amazon Bedrock AgentCore on 5 November 2026. Web Search hands your agent ranked links and snippets and leaves synthesis to your own model. Web Answer runs the retrieval loop, model, synthesis and citation step and returns a finished cited answer through AgentCore Gateway. This piece compares the returned artifact, the permissions, the controls you keep, the $10 per 1,000 tool calls price and the limits, including what AWS's guide does not establish.
AWSAgentCoreWeb search
Nov 11, 2026
AgentCore memory poisoning: what AWS builds in, what its new write-validation pattern adds, and where the demo evidence stops
An AWS Security Blog post from 9 November 2026 separates the memory-poisoning controls AgentCore already offers (IAM namespace conditions, namespace-variable validation, Cedar policy, custom consolidation prompts, the ApplyGuardrail API) from a custom pre-write Lambda interceptor that scores each write and returns ALLOW, QUARANTINE or DENY. The custom layer is a sample you deploy, not a default setting. Its published 11-scenario demo used a keyword-overlap stand-in for embeddings, so the results do not show production accuracy or false-positive rates. Writes that bypass AgentCore Gateway, such as direct SDK calls, are not covered by the interceptor.
AWSAgentCoreAI security
Nov 11, 2026
Bedrock's AgenticRetrieveStream now takes maxAgentIteration 0 or 1: passages only, or one retrieval plus one answer
On November 9, 2026, AWS said Amazon Bedrock Managed Knowledge Base can run retrieval-only and single-shot RAG requests through AgenticRetrieveStream. Setting maxAgentIteration to 0 with generateResponse false returns passages with no foundation model call. Setting it to 1 retrieves once and generates one answer with no planning rounds. Values 2 to 5 keep the existing agentic behavior. The application chooses the mode; the parameter does not route requests automatically. AWS gives no new latency, cost or accuracy measurements, and the passages-only stream differs in shape from a direct Retrieve call.
AWSAmazon BedrockRAG