Arc Brief · Methodology

How Arc Brief Works

How we find, generate, review, and publish our daily APAC technology intelligence briefings.


What Arc Brief Is

Arc Brief is an independent APAC-focused technology and infrastructure intelligence publication. We publish one curated intelligence briefing per category per day — six categories in total — covering data centres, AI adoption, technology policy, investment, and the impact of technology on society across Australia, Singapore, India, Malaysia, Japan, and Southeast Asia.

Arc Brief is written for executives, investors, and policy professionals who need to understand what is happening in APAC technology without reading twelve newsletters a day.

How Stories Are Generated

Each daily briefing is generated using Claude, Anthropic’s AI model, with live web search enabled. The model is instructed to find one real, current, significant news story per category, prioritising stories that are primarily about or directly relevant to APAC markets.

The generation prompt includes:

  • A requirement to lead with APAC consequences, not global context
  • Category-specific search guidance and story-size hierarchy (e.g. hyperscaler capital commitments above $1B take priority over smaller announcements in the Data Centers category)
  • A deduplication instruction to avoid covering the same story across two categories
  • An instruction to prefer open-question framing for contested or emerging topics, and declarative framing for straight announcements

Each story is returned as structured JSON including a headline, subtitle, body (three paragraphs), APAC-specific analysis, country-level angles for Australia, Singapore, and India, and an SEO title and description.

DISCLOSURE: Arc Brief’s daily briefings are AI-generated using live web search. They are not written by a human journalist in the traditional sense. The prompts, category guidance, and editorial standards are written and maintained by S J Okafor.

Editorial Standards

Stories must be about a real, verifiable event. The model is instructed to search for current news and cite sources. Stories that cannot be grounded in a real, recent event should not be generated — the model is penalised in its instructions for fabricating events.

We apply the following editorial standards to every briefing:

  • APAC primacy. US and European stories appear only when they carry direct, named consequences for APAC markets. Even then, the APAC angle must be the lead.
  • Significance hierarchy. Within each category, we apply a story-size hierarchy. In Data Centers, for example, hyperscaler capital commitments above $1B always take priority over smaller announcements.
  • No wire copy framing. Headlines must lead with the most interesting consequence, tension, or number in the story — not with the company name or the action verb. Banned words in headlines include “announces,” “unveils,” “reveals,” and “launches.”
  • Question framing for contested topics. If a story involves an open question, a contested claim, or an emerging trend, the headline is written as a question. Straight announcements use declarative headlines.

Significance Scoring

Every generated article receives a significance score (0–10) computed at generation time based on the content of the article. The scoring criteria are:

  • +3 if the story involves an investment or deal value above $1B
  • +2 if the story involves a named hyperscaler or major operator (Microsoft, Amazon, AWS, Google, Meta, Blackstone, AirTrunk)
  • +2 if the story directly involves Australia or Singapore
  • +1 if the story involves India or another APAC market (Japan, Malaysia, Indonesia, Vietnam, Thailand)
  • +1 if the story involves a policy decision or regulatory action affecting multiple markets

On the homepage, the article with the highest significance score is promoted to the lead position when viewing all categories. Within a single-category view, the most recent article leads.

Editorial Content

The long-form editorial series — Issues 001 through 006 and counting — is written and researched by S J Okafor. These are not AI-generated. They represent original analysis of data infrastructure, technology governance, and the forces shaping digital Asia, drawing on primary data, industry reports, and independent reporting.

Corrections

If you identify an error in a daily briefing or editorial, please contact us at the address listed below. AI-generated briefings may contain factual errors despite the safeguards above. We take corrections seriously and will update content where errors are confirmed.

Contact: editor@arcbrief.news

APAC tech, read once a week.

No ads, no noise.

Subscribe →