Company/The Signal

    THE

    SIGNAL

    Curated intelligence on AI, enterprise technology, and the systems reshaping how businesses operate.

    0 dispatches
    Data & Infrastructure
    Data & Infrastructure◆ Latest

    3 October 2026 · 5 min read

    Always-On Agents Turn the AI Question Into an Infrastructure Question

    OpenAI's always-on Dots agents, its AWS-hosted agent option and the open-source OpenClaw Enterprise control plane all landed in one week. Together they move the enterprise AI decision from which model to use to where agents run and who controls them.

    By En Interactive

    Read full dispatch →
    Data & Infrastructure
    Data & Infrastructure

    3 October 2026

    Always-On Agents Turn the AI Question Into an Infrastructure Question

    OpenAI's always-on Dots agents, its AWS-hosted agent option and the open-source OpenClaw Enterprise control plane all landed in one week. Together they move the enterprise AI decision from which model to use to where agents run and who controls them.

    5 min readEn Interactive
    Industry News
    Industry News

    26 September 2026

    The UN's Agent Warning Raises a Simpler Question: Who Watches Yours?

    A UN-backed scientific panel says the safeguards around AI agents are unravelling. For most businesses, the urgent question is not loss of control but whether anyone is counting and reviewing what their agents actually do.

    5 min readEn Interactive
    Agentic Systems
    Agentic Systems

    18 September 2026

    Salesforce's Named Agents Signal a Shift From AI Features to Roles

    Salesforce shipped seven named AI agents this week -- Casey, Paige, Carter, Hunter, Marshall, Piper, and Fin -- each mapped to a specific business function rather than sold as general AI capability. For enterprise buyers, that reframes agentic AI from a technology decision into a staffing decision, with real implications for how these systems get evaluated, budgeted, and governed.

    5 min readEn Interactive
    AI Models
    AI Models

    11 September 2026

    Four AI Labs, One Week: The Real Cost of Model Release Fatigue

    Anthropic, Google, Meta, and OpenAI each shipped new model updates within the same week, and no single lab leads both reasoning and coding benchmarks. For enterprise teams, the real cost isn't picking the wrong model — it's the procurement cycle triggered every time a new one ships.

    5 min readEn Interactive
    Enterprise Tech
    Enterprise Tech

    4 September 2026

    Anthropic's Enterprise Frontier Safeguards Address AI's Data Custody Deadlock

    Anthropic's new Enterprise Frontier Safeguards let companies run Claude under zero data retention while misuse-monitoring logs stay inside their own cloud account instead of Anthropic's. The move targets the compliance objection that has stalled enterprise AI deployment more consistently than any capability gap.

    5 min readEn Interactive
    Agentic Systems
    Agentic Systems

    4 September 2026

    Google's Gemini Enterprise for Legal Targets the Cost of Contract Review

    Google Cloud launched Gemini Enterprise for Legal on August 25, a suite of AI agents for contract review, citation verification, and regulatory monitoring, with law firms Cleary Gottlieb and Weil Gotshal already on board. The launch matters less as legal technology than as a preview of what happens when hourly-billed document work becomes a configurable platform skill.

    5 min readEn Interactive
    Agentic Systems
    Agentic Systems

    28 August 2026

    OpenAI Presence Has No Self-Serve Tier — That Is the Real Story

    OpenAI launched Presence this week — an enterprise platform for AI voice and chat agents that is explicitly not self-service. The design choices behind Presence reveal more about the true state of production AI agent deployment than any benchmark result.

    5 min readEn Interactive
    AI Models
    AI Models

    14 August 2026

    DeepSeek's V4 Pro Benchmarks Look Impressive. Nobody Has Verified Them.

    DeepSeek's V4 Pro model reached general availability on August 12 with vendor-reported benchmark gains as high as 49.9 points, but no independent evaluator has reproduced the numbers, and a NIST evaluation found a serious safety gap in the same model line. Five days later, DeepSeek confirmed API prices will rise as much as 12-fold starting August 17, changing the cost calculus that made the model attractive in the first place.

    5 min readEn Interactive
    AI Models
    AI Models

    12 August 2026

    What Anthropic's Custom Silicon Bet Means for Enterprise AI Cost

    Anthropic confirmed this week it is assembling an in-house team to design custom chips for Claude. For enterprise teams managing inference costs today, the implication is a three-to-five-year story — not a quarterly one.

    4 min readEn Interactive

    Stay current

    The Signal, direct to your inbox.

    A few times a month. Only what matters. No noise.