Claude Managed AgentsAnthropic Requires Safety Patches Within One Business Day
Anthropic's self-hosted Claude Managed Agents beta keeps sessions and credentials in your network and is the only Managed Agents option eligible for ZDR. The licence requires safety updates within 1 business day and rollbacks within 8 hours.
November 14, 2026 · 10 min readFireworks AIRotate Every Secret You Stored on Fireworks After Breach
Fireworks AI's November 9 update widens its October breach to billing data, stored secrets, datasets and incomplete model weights. Rotate every secret you stored there and shrink what you leave at training vendors.
November 10, 2026 · 10 min readClaude Fable 5.2Claude Fable 5.2 Requires 30-Day Retention for Most Buyers
Claude Fable 5.2 costs $8/$40 per million tokens, twice Opus 5.5, and most buyers must accept 30-day retention. Anthropic still recommends starting with Opus 5.5, so test effort levels before you change a default.
November 9, 2026 · 11 min readzero data retentionUse Bedrock's None Mode to Enforce Zero Data Retention
A zero data retention agreement covers specific endpoints and models. OpenAI and Anthropic leave out files and batches, and Fable requires 30-day retention, so approve ZDR per endpoint.
November 9, 2026 · 15 min readbatch record reviewEnveda's Claude Reviewer Cut Review Time From 1,553 to 166 Minutes
Enveda's Claude-based reviewer caught all 11 human-found issues in 1,061 pages of executed batch records, raised 7 more and cut review time from 1,553 to 166 minutes. It stays advisory-only, outside GMP validation scope, and the study covers four records.
October 27, 2026 · 10 min readOpenAI on Google CloudGoogle Cloud Lets Codex Users Run OpenAI Models on Its Platform
GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna now draw down Google Cloud commits, but Google's savings-plan terms give third-party models no discount, and the private preview has no published price, regions or ZDR yet.
October 22, 2026 · 10 min readopen-weight inferenceTogether vs Fireworks vs Groq: Fireworks Leads Moonshot's Audit
Kimi K3 costs $3 / $15 at four of the five hosts compared that offer it and gpt-oss-120b $0.15 / $0.60 at four of six, but Moonshot's vendor audit found schema accuracy from 100% down to 71.96%. Fireworks is the default for agents; Together is the one to avoid for tool-heavy work.
October 17, 2026 · 15 min readweb search APIExa Stays Near the Top in Tavily vs Exa vs Brave Benchmarks
Exa is the safest default web search API for AI agents, and Tavily is the one to skip. At frontier-model prices, the page text an API returns can cost 6 to 20 times the search fee.
October 16, 2026 · 15 min readOpenAI Decisions APILong Rubrics Can Make OpenAI Decisions API Cost 4x More
OpenAI's Decisions API bills only input tokens at $0.10 per 1M, which saves a little on short tickets. It also drops Luna's cached-input discount, so long shared rubrics can cost about four times more than the Responses API.
October 8, 2026 · 9 min readClaude for Google WorkspaceClaude Workspace Add-On Does Not Support Zero Data Retention
Anthropic's Claude add-on for Google Docs, Sheets and Slides deletes data within 30 days regardless of your retention setting, has no ZDR or Bedrock/Vertex routing, and is barred from PHI in beta. Allowlist it for a pilot group.
October 7, 2026 · 10 min readCyber Verification ProgramAnthropic Adds Red Team Access to Cyber Verification Program
Anthropic's Cyber Verification Program now has Defense, Red Team and Specialized tiers. Each requires data retention for misuse monitoring, so security leaders must decide whether SOC and pen-test data can sit in a vendor's logs.
October 6, 2026 · 9 min readCloudflare AI GatewayCloudflare's Auto Router Fails 4x as Often as Opus on Its Own Test
Cloudflare's Auto Router solved 86.6% of its own benchmark tasks against Opus 5.5's 96.6%, at 40% of the cost per success. The saving only holds where a failed task costs under about 13 cents.
October 1, 2026 · 11 min readCursor alternativesCursor Alternatives That Survive Security Review Start With Copilot
Cursor cannot be self-hosted and every request is built on its servers. Standardise on Copilot Business, run Claude Code through your own cloud account, and reserve Tabnine or Cline for air-gapped shops.
September 30, 2026 · 13 min readAI vendor security reviewAI Vendor Security Review: 6 Questions That Change the Answer
Six questions decide an AI vendor security review, and a SOC 2 report answers none of them. How OpenAI, Anthropic, Microsoft, Google and Amazon Bedrock answer each, as of September 2026.
September 12, 2026 · 20 min readCursorOpenAI Cuts Cursor Off Nov 12. BYOK Voids Your ZDR.
OpenAI stops serving its models to Cursor on 12 November 2026. Pointing Cursor at your own OpenAI key is not a swap — Cursor's documentation says its Zero Data Retention policy does not apply to custom keys, and every request still routes through Cursor's backend.
August 30, 2026 · 12 min readzero data retentionFable 5 Needs Retention On. ZDR Was Never Zero.
Claude Fable 5 and Mythos 5 are Covered Models: they refuse to run in a zero-data-retention workspace on the Claude API, Bedrock, Google Cloud or Microsoft Foundry. Anthropic's own docs also say flagged prompts can be held up to two years even under ZDR — while OpenAI previews the opposite bet.
August 20, 2026 · 13 min readOpenRouterStripe Bought OpenRouter. A Toggle Is Not a Contract.
Stripe is reportedly paying over $7 billion for the gateway 8 million developers route prompts through. Its privacy toggles bind the downstream model providers, not OpenRouter itself — and the only DPA anybody countersigns is enterprise-tier.
August 17, 2026 · 13 min read