Google introduces a persistent work agent; Haiku 5.5 targets everyday AI tasks
Google announced a unified Gemini agent at Gemini at Work on October 8. The company says it can research, create content, write and run code, and connect to business systems from one interface. Cloud execution lets longer tasks continue after a laptop closes, while the agent retains context across devices. Google also describes enterprise identities, permissions, audit trails and spending controls. This is an announcement of the agent’s capabilities; access and rollout should be checked for each organization.
Also from October 7: Anthropic introduced Claude Haiku 5.5 for frequent, narrowly scoped work such as classification, summaries and database queries. It can also act as a coding subagent alongside larger Claude models. Haiku now offers adjustable effort, allowing developers to balance response quality and cost. Anthropic reports lower pricing than Haiku 4.5, with rates that depend on prompt length. Its release page lists availability on the Claude Platform and major cloud providers.
A learning question from October 7: Google Research reported a three-month randomized study of 133 patent lawyers at 11 firms. AI assistance improved patent drafting, but gains in independent judgment differed by experience. Senior lawyers with AI access performed better on a later task without AI; junior lawyers showed no average improvement on that task. The researchers caution that the sample and observation period are limited. The practical lesson is to assess independent skills separately from the quality of AI-assisted work, rather than assuming one proves the other.