AI News — July 2, 2026

Model releases

• [2026-07-01] Anthropic — Fable 5 and Mythos 5 are back after the US lifted export controls. On June 30 the Commerce Department lifted the June 12 export-control directive that had forced Anthropic to suspend both frontier models globally, ending a 19-day blackout; Anthropic said it received notice and would begin restoring access the next day. Fable 5 returned July 1 across Claude.ai, the Claude Platform, Claude Code, and Claude Cowork, with AWS, Google Cloud, and Microsoft Foundry re-enablement to follow “as soon as possible.” It comes back as a battened-down build behind a new cybersecurity safety classifier — trained in coordination with the government — that targets and blocks the safeguard-bypass behavior (a jailbreak Amazon researchers surfaced) that triggered the original order. Access terms are temporary: Pro, Max, Team, and select Enterprise plans get Fable 5 at up to 50% of weekly usage limits through July 7, after which it’s available via usage credits. It matters because the last frontier piece is restored — and unlike Mythos 5’s June 26 org-gated return, Fable 5 comes back as a general unlock, but with a new model-level cyber classifier now part of the cost of shipping it. (official, CNBC)

  Anyone who rewired production off Fable 5 during the 19-day gap gets it back on a clock: the 50%-of-limit cap and the July 7 cutover to usage credits make this a return to promotional access, not restored headroom, so the switch-back is worth planning around the credit terms rather than treating July 1 as business-as-usual.

• [2026-06-30] Anthropic — Claude Sonnet 5 shipped, a new mid-tier model with a native 1M-token context window, positioned as “the most agentic Sonnet yet” and close to Opus 4.8 quality at a lower price. Introductory API pricing is $2 / $10 per million input/output tokens through August 31, then $3 / $15. It’s the default model on Free and Pro (available to Max, Team, and Enterprise) and is already the default in Claude Code (v2.1.197). On Anthropic’s launch benchmarks it beats Sonnet 4.6 across the board (reported SWE-bench Pro of ~63% vs ~58%), landing between Sonnet 4.6 and Opus 4.8 (~69%) on agentic coding. It matters because it widens the cost-performance menu — near-frontier agentic behavior at a mid-tier price, with a 1M context Sonnet didn’t previously have. (official)

  Sonnet gaining a native 1M-token window is the quieter half of the release: long-context work that used to force a top-tier model just to fit the tokens can now run at $2/$10, so the question of what actually needs Opus narrows to genuine reasoning depth rather than context size.

  For Software Developers: Point a Claude Code session at a large multi-file change — a whole service plus its tests and config — without pre-trimming to fit the window, since Sonnet 5’s native 1M context holds it, and do it at $2/$10 through August 31 instead of reaching for Opus just to fit the tokens.

Coding agents

• [2026-07-01] Anthropic — Claude Code v2.1.197–v2.1.198 shipped. v2.1.197 (June 30) makes Sonnet 5 the default model with a native 1M-token context. v2.1.198 (July 1) brings Claude in Chrome to general availability, adds background-agent notifications via a new Notification hook (agent_needs_input / agent_completed), and — the substantive change — lets background agents auto-commit, push, and open a draft PR on completion; it also adds Claude Platform on AWS as a gateway upstream and a /dataviz skill for chart design. It matters because the background-agent auto-PR flow closes the loop for fire-and-forget runs — the agent lands reviewable work without a human shepherding the git steps — while Sonnet 5 as the default resets the price/quality baseline for everyday sessions. (official)

  Auto-commit-and-open-a-draft-PR turns a background agent from something you have to babysit through the git steps into a fire-and-forget worker — which also moves the control point to reviewing that draft PR, since the agent’s output now lands in branch history without a human running the push.

  For Software Developers: Hand a background agent a scoped chore — bump a dependency and fix the fallout — before you step away; with v2.1.198 it commits, pushes, and opens a draft PR on its own, and the new agent_completed notification pings you when there’s a PR waiting to review instead of a terminal to check.

Watch list

• GPT-5.6 general availability — Sol, Terra, and Luna remain a ~20-partner government-coordinated limited preview; OpenAI says GA is “in the coming weeks” (mid-July per Axios reporting). Watch for the GA date, whether the vetted-partner gate persists, and published third-party benchmarks beyond the internal CTF cyber scores. (prior coverage)

  What would make this significant is the gate itself: if GA keeps the ~20-partner vetting rather than opening broadly, it signals OpenAI settling into the same access-controlled posture the Fable 5 saga forced on Anthropic, not a clean public launch.

• Gemini 3.5 Pro (GA) — July has arrived with the model still a limited Vertex AI enterprise preview: no model card, public benchmarks, or pricing for the promised 2M-token context and Deep Think mode. Pichai’s Google I/O “give us until next month” has now lapsed a second time. Watch for a July rate card or a slip into Q3. (prior coverage)

  A July rate card would matter beyond punctuality: the promised 2M-token context is the one spec that would leapfrog Sonnet 5’s new 1M window, so each slip leaves that long-context tier uncontested a little longer.

• Fable 5 usage terms (July 7) — Fable 5 is back but on temporary terms: 50% of weekly usage limits on paid plans through July 7, then usage-credits only. Watch how consumer and enterprise access settle once the promotional window closes, and whether the new cybersecurity classifier over-blocks legitimate security work. (official)

  July 7 is the first read on what Fable 5 actually costs to run at steady state rather than during a promo, and the first chance to see whether the classifier’s safety gain carries a false-positive tax on legitimate security and red-team work.

• Alibaba / Qwen distillation amendment — the Hagerty–Kim amendment to must-pass defense legislation (which would blacklist or sanction Chinese firms found improperly accessing US AI-model output) floated for June 24 still shows no confirmed filing. Watch for whether it’s formally entered into the bill’s text and Alibaba’s still-absent response. (prior coverage)

  An actual filing would mark a change in tactic — from gating API access toward sanctioning the firms accused of distilling US model outputs in the first place — which is the enforcement lever left once export controls prove unable to hold a capability.

• MCP spec finalization (July 28) — the 2026-07-28 release candidate (stateless core, Extensions framework, Tasks, MCP Apps, authorization hardening, formal deprecation policy) is in its validation window, with the Ruby/TypeScript/Python SDKs updating against it. Watch for the rest of the SDK support landing — and any breaking issues in Tasks, MCP Apps, or the new MCP-specific HTTP headers — before the cutover. (official, prior coverage)

  What makes 2026-07-28 the one to watch is the formal deprecation policy it introduces: this becomes the release that fixes the compatibility contract everything built afterward has to honor, with the stateless core deciding whether MCP servers can scale horizontally like ordinary web services.