The Frontier
5 left

Agents & software

Coding agents, tools, RAG, orchestration — what you can run this week vs a demo.

Agents & softwareDispatch

Apple opens Siri AI to developers and withholds it from EU iPhone

8 June 2026 WWDC: developer testing now, user beta later this year. DMA update: no Siri AI on iOS 27, iPadOS 27, or paired watchOS in the EU. macOS 27 and visionOS 27 still get it.

TLDR

Apple's 8 June 2026 newsroom post introduces Siri AI, powered by Apple Intelligence, with developer testing starting that day and a user beta later in 2026. A same-day DMA update says Apple will not ship Siri AI in the European Union with iOS 27 and iPadOS 27, and therefore not on watchOS 27 for EU users who need a paired iPhone. EU users will get Siri AI on macOS 27 and visionOS 27. EU-located developers cannot test the new Siri AI features for apps on iOS, iPadOS, and watchOS 27. Consumer ship month is UNKNOWN.

Agents & softwareGround Truth

JSON Schema makes agent tools production-ready

NIST AI 800-5 finds commenters agree agent threats are novel and existing controls need adaptation. Schema syntax is not the bar.

TLDR

Claim: valid JSON against a schema is enough to ship side-effecting agent tools. NIST Trustworthy and Responsible AI 800-5 (May 18, 2026) summarizes CAISI's agent-security RFI: commenters widely agreed agents present novel threats and fundamental cybersecurity practices require adaptation. The January 2026 RFI foregrounded indirect prompt injection and misaligned objectives. Verdict: overstated for workflows with side effects.

Agents & softwareField Notes

Vertex Gen AI eval pipeline replaces RAG demo scorecards

Google’s Vertex evaluation service and EvalTask score retrieval and generation separately; model-based rubrics replace hallway comparisons.

TLDR

Google Cloud documents a Gen AI evaluation service on Vertex AI: generate answers for a prompt set, score them with named rubrics, and log runs in Vertex AI Experiments. For RAG, operators use EvalTask datasets with prompt and response columns, then batch jobs for large golden sets. Groundedness and answer-quality metrics replace anecdotal pilots. Named customer before/after scores are UNKNOWN.