Anthropic Weekly Digest — June 29-July 5, 2026

This was a repair-and-expansion week for Anthropic: the company pushed a new Sonnet-class model into default use, opened a science workbench beta, restored access to Fable 5 after U.S. export controls were lifted, and signed a California public-sector deal.

Anthropic introduced Claude Sonnet 5 on June 30, positioning it as its most agentic Sonnet model to date and making it the default model for Free and Pro users while also making it available to Max, Team, Enterprise, Claude Code, and the Claude Platform users.1 The pricing move matters for builders: Sonnet 5 launched at $2 per million input tokens and $10 per million output tokens through August 31, 2026, before moving to $3 and $15 respectively.1

The model is meant to narrow the gap between Anthropic's mid-tier Sonnet line and the more expensive Opus tier. Anthropic says Sonnet 5 gets closer to Opus 4.8 on agentic work such as coding, tool use, and professional tasks, while keeping a lower price profile.1 The safety section is just as important as the performance pitch: Anthropic says Sonnet 5 showed a lower overall rate of undesirable behaviors than Sonnet 4.6, but a higher rate than Opus 4.8 and Claude Mythos Preview on its automated behavioral audit.1

Anthropic also launched Claude Science in beta on June 30 for Pro, Max, Team, and Enterprise users.2 The product combines a coordinating agent, more than 60 curated scientific skills and connectors, specialist agents, and a reviewer agent that checks citations and calculations.2 It runs on macOS, Linux, remote machines over SSH, and HPC login nodes, which keeps large or sensitive datasets on the systems where researchers already work.2

The launch came with a grant-style adoption push. Anthropic said it will support up to 50 Claude Science AI for Science projects with up to $30,000 in credits, while Modal will provide up to $2,000 in compute for select projects; applications are open through July 15, notifications are due July 31, and projects run from September 1 to December 1.2 CNBC added one new strategic detail from the launch event: Anthropic is starting an internal drug-discovery program focused on neglected diseases, with life sciences head Eric Kauderer-Abrams saying the company wants tighter feedback loops with the industry it is selling into.3

The Fable 5 shutdown moved into a negotiated restart. Anthropic said on June 30 that U.S. export controls on Claude Fable 5 and Claude Mythos 5 had been lifted, and that Fable 5 would return globally on July 1 across Claude Platform, Claude.ai, Claude Code, and Claude Cowork.4 The company also said Pro, Max, Team, and select Enterprise plans would include Fable 5 for up to 50% of weekly usage limits through July 7, after which access would move to usage credits.4

The restart did not erase the underlying policy fight. Anthropic said the June 12 export-control order followed a report from Amazon researchers about bypassing Fable 5 safeguards, and that a new classifier now blocks the specific reported technique in more than 99% of cases.4 The company also said researchers from the U.S. Department of Commerce's Center for AI Standards and Innovation tested its prior and new safeguards.4

On July 2, Anthropic published a more detailed framework for Fable 5 cyber safeguards and jailbreak severity.5 The proposed Cyber Jailbreak Severity scale runs from CJS-0 to CJS-4 and scores jailbreaks across capability gain, breadth of capability gain, ease of weaponization, and discoverability.5 Anthropic also opened a HackerOne program for Fable 5 cyber-jailbreak submissions and invited feedback at a dedicated cyber-safeguards address.5

External reporting broadly matched the company's account. The Guardian reported on July 1 that Anthropic had restored access after more than two weeks of blackout and that Commerce Secretary Howard Lutnick said the government had worked with Anthropic to analyze and approve Fable 5.6 CBS News separately reported that Anthropic said it would begin restoring access Wednesday after the federal government lifted restrictions.7

California announced a statewide Anthropic partnership on June 29, making Claude the first AI productivity tool available to all state agencies through the California Department of Technology's Statewide Information Technology Shared Services portal.8 The agreement gives state agencies, cities, and counties access to Claude at a 50% discount, plus free workforce training, technical assistance, and workflow input from Anthropic developers.8

The governor's office listed several existing California uses: Engaged California, Poppy, cyber defense work by the Department of Technology and CalOES, DMV customer-service improvements, and internal workflows at the Department of Health Care Services.8 The Los Angeles Times reported the same discount terms and quoted Newsom saying AI should help government workers move faster rather than replace their work.9

Funding/valuation, leadership, new lawsuit filings, standalone infrastructure, and compute-supply had no confirmed material entries for this window in readable public sources. The China-access policy story reported by the Financial Times was visible only behind a subscription barrier, so it remains an unconfirmed lead rather than a digest entry.10