How the answers frame IHG
This snapshot was taken on 2026-09-29. It covers 250 unbranded hotel and travel questions put to ChatGPT, Claude, Gemini, Perplexity and Grok, or 1,250 answers in all. IHG was named in 428 of them. Marriott (687) and Hilton (681) were named more often, and so was Hyatt (539). Grok named IHG most (138 of 250 answers), followed by ChatGPT (100), Gemini (75), Claude (58) and Perplexity (57).
In the stored answers, IHG most often arrives as the closing name in a group. Claude's answer on service consistency for business travelers listed Marriott Bonvoy, Hilton Honors, Hyatt and IHG properties. Perplexity's answer on front desk employers opened with 'Hilton, Marriott, Hyatt, and IHG'. Its IHG entry offered little detail beyond a suggestion to include it in the search, and it noted that individual hotel management makes a big difference. Most credit tags on the sampled answers were 'Not Substantive', meaning IHG was named without a specific reason attached.
Some questions did give IHG a distinct role. Visibility by question intent was highest for loyalty and perks (62%), then sustainability (46%) and safety (40%). It was lowest for 'avoid' questions (16%), trust (23%) and service (24%). Gemini's technology answer tied IHG to AI-managed energy use, and its investing answer described IHG as ranking among the top companies by profit on a franchising model. Gemini also listed Staybridge Suites for families, and Grok listed Holiday Inn Express. The IHG One Rewards Premier card appeared in Gemini's status answer.
One answer put IHG in a less comfortable frame. Asked which hotel loyalty programs have been hacked, Claude listed IHG One Rewards alongside Marriott and Hilton. It gave IHG one sentence and no incident date or company response.
IHG One Rewards has been subjected to data breaches that affected millions of consumers.
What the press is saying
The PeakMetrics news index shows 3,885 IHG mentions in the last 30 days, with the brand in the headline of 737. Recent headlines fall into three groups. The first is consumer promotions, such as a 30% destination deals offer for stays through November 8, 2026. The second is corporate finance, including a euro bond deal and an own-shares transaction notice. The third is expansion, with the Noted Collection brand debuting in Japan with two Tokyo hotels.
That volume is well below Marriott (25,298), Hilton (14,335) and Hyatt (14,627). It is also below Airbnb (16,728) and Expedia (8,663).
Often cited for its commitment to sustainability-focused technology, using AI to manage energy consumption (HVAC and lighting) based on real-time occupancy data.
The gap
The press story is about deals, capital markets and a new brand launch. The AI story is about where IHG sits in a category shortlist and which programme features it is associated with. The Noted Collection launch and the promotional headlines did not show up in the sampled answers.
Press volume and AI visibility also do not move in step in this snapshot. IHG has less news than Airbnb, Expedia and Wyndham, yet the assistants named it far more often than any of them (428 answers, against 158, 297 and 234). Airbnb has the fourth-largest news count in this peer set and one of the smallest AI footprints. Coverage alone does not predict who the assistants reach for.
IHG consistently ranks among the top companies by profit, supported by a resilient franchising model.
What to watch
Three points for an IHG comms team, all from one snapshot rather than a trend. First, being on the list is not the same as owning a storyline. Many mentions carried no specific credit, so the questions where IHG gets a distinct role (loyalty, sustainability, safety, family suites) are worth mapping against the sources feeding those answers.
Second, the data-breach line in Claude's answer is a reputational exposure that the news headlines do not reflect. It is worth checking how often assistants raise it, and whether answers give any company response.
Third, the five assistants differ sharply in how often they name IHG. Grok named it about twice as often as Claude or Perplexity, so a single-assistant check would give a misleading picture. Live answers vary between runs, so repeat the benchmark questions before drawing conclusions.