Anthropic settled Fable 5’s month-long billing whiplash on July 17: starting July 20, Max and Team Premium get Fable 5 permanently at 50% of a usage pool that’s also shrinking, while Pro and Team Standard get a one-time $100 credit before falling to $10/$50-per-million-token metering. The reversal came days after GPT-5.6 Sol and Moonshot’s 2.8-trillion-parameter Kimi K3 both undercut Fable 5 on price.
Atlassian launched Jira Planner on July 15 — a spec-generation tool that pulls codebase, Jira, and Confluence history into a structured technical spec before any coding agent touches a task. Atlassian’s own numbers: AI usage among engineering teams is up 65%, but velocity gains have plateaued around 15%, because coding is only 15-16% of the SDLC. Claude Code and Cursor are now GA as first-class agents inside Jira; Codex is still ‘coming soon.’
Four days after being caught silently uploading entire git repositories to a Google Cloud bucket, xAI open-sourced Grok Build’s 844,530-line Rust codebase under Apache 2.0 on July 15 — but the upload function is still compiled into the binary, disabled only by a server-side flag xAI can flip back on without shipping an update. External contributions are rejected and GitHub issues are disabled.
Bloomberg reported July 16 that Gemini 3.5 Pro is months behind its promised June ship date because a late-June retrain aimed specifically at improving coding performance produced disappointing results — the first specific, on-the-record reason given for a delay this blog has tracked since a June 30 DeepMind talent-exodus story. Alphabet shares fell as much as 4.4%, erasing roughly $200 billion in market value, the same week rivals shipped competitive coding models.
Anaconda has acquired Kilo Code, an open-source, model-agnostic agentic coding platform used by 3 million developers and processing nearly 10 trillion tokens a month, to build an enterprise cost-and-governance layer on top of 500+ AI models. It’s the third major agentic coding platform swallowed by a non-AI-native parent in 2026, and it bets on brokering across models rather than owning the stack the way Claude Code does.
Independent wire-level traffic analysis published July 13-14 found xAI’s Grok Build CLI silently uploading entire git repositories — full history, unredacted secrets, files the agent never even opened — to a Google Cloud Storage bucket. On a 12GB test repo, model traffic was ~192KB while the upload channel moved 5.10GiB: a roughly 27,800x gap between what the task needed and what left the machine. The ‘Improve the model’ privacy toggle didn’t stop it; only a quiet server-side flag flip did.
Wiz Research disclosed GhostApproval on July 8: a symlink trust-boundary flaw in six AI coding agents — including Claude Code, Cursor, Windsurf, Amazon Q, Augment, and Google Antigravity — where the agent’s own reasoning correctly flags a malicious file target, but the permission dialog shown to the user displays the decoy filename instead. Anthropic initially rejected the report as outside its threat model; the fix that quietly landed nine days earlier turned out to be unrelated.
Meta launched Muse Spark 1.1 on July 9 — its first monetized model API, priced at $1.25/$4.25 per million tokens with a 1M-token context window. Meta’s self-reported Terminal-Bench 2.1 score of 80.0 didn’t survive contact with Vals AI’s independent rerun, which measured 69.29 — a gap large enough that a Hacker News commenter alleges the benchmark’s resource limits were exceeded.
Anthropic extended free Fable 5 access for Pro, Max, Team, and premium Enterprise seats a second time, from July 12 to July 19, 2026 at 11:59:59 PM PT — pushing the switch to $10/$50-per-million-token metered billing back to July 20. It’s the second last-minute reprieve in six days, on a model that was already offline for 19 days this quarter over export controls.
Artificial Analysis ran Grok 4.5 through its own independent harness rather than trusting Cursor’s launch numbers. The verdict: genuinely frontier-class intelligence (4th place, behind Fable 5, GPT-5.5, and Opus 4.8) and a real token-efficiency edge — bought at the cost of a hallucination rate that more than doubled, from 25% to 54%.