Position
Control of the battleground's chokepoint: customers, distribution, standards.
Scored 1–5AI POWER RANKINGS · EDITORIAL STANDARD
AI Power Rankings: Method
Every placement is an editorial assessment grounded in public evidence. Scores are built from three dimensions, then grouped by weighted score.
Three dimensions
Scale
Market cap or latest private valuation, in bands.
Scored 1–5Momentum
Recent moves.
Scored 1–5Momentum scoring rules
- Count only moves from the last 90 days that the company itself made in that battleground, with materiality 2 or more, verified or reported by a named source.
- Exclude regulatory actions, cancellations, and events that happen to a company rather than moves it makes.
- Acquisitions, capital raises and IPOs, and committed partnerships count 2 points; other moves count 1.
01
1–22
3–43
5–64
7+5
Points → score
Tier cutoffs
- Leader
- 4.0 and above.
- Contender
- 3.0–3.9.
- Challenger
- 2.0–2.9.
- Watch
- below 2.0.
Scale bands
Frontier models
- 5 = over $500B, 4 = $100–500B, 3 = $10–100B, 2 = $1–10B or undisclosed.
Compute and cost
- 5 = over $1T, 4 = $200B–1T, 3 = $30–200B, 2 = $5–30B.
EVIDENCE STANDARD
Confidence
Companies with thin or single-source evidence are marked Low confidence and cannot rank above Challenger.
ACCOUNTABILITY
Editorial rules
- Tiers are RAW PEC's assessment, not fact.
- Every placement cites the public moves behind it.
- RAW PEC does not accept payment, sponsorship or event fees tied to a company's tier or coverage.
- Disclosure: RAW PEC advises companies in the ServiceNow ecosystem; placements involving ServiceNow, or any company RAW PEC works with, get a second review before publishing.
- Companies that dispute a placement with evidence get a review within one cycle, and changes are logged with the reason.
PUBLICATION CADENCE
Re-scored monthly
Full re-rank monthly, in the first week. Mid-month tier changes happen only after a major event (an acquisition, IPO or committed partnership).
Back to AI Power Rankings