Should my current SEO agency already handle AI search?

A competent SEO agency is already doing most AI search work. Google states that to appear as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible for a snippet, and that there are no additional technical requirements. That is core SEO.

The gaps sit outside Google. ChatGPT, Claude and Perplexity run their own crawlers, with their own rules and their own failure points. An agency that only checks Googlebot can miss all of them.

So you probably don't need a second agency. You need to confirm the current scope covers three things: per-bot access, firewall testing and fact checks in AI answers. Our guide to what AI SEO services should include lays out the full list.

These tasks serve classic search and AI search at the same time:

• Indexing. A page Google has not indexed cannot be cited in its AI features.
• Rendering. Vercel found that ChatGPT and Claude crawlers fetch JavaScript files but do not execute them, so content that appears only after JavaScript runs may be invisible to them. Googlebot renders, which is why an agency testing only in Google may never see the problem. Ask for a raw HTML check of pricing and key product facts.
• robots.txt. Same file, more user agents to manage.
• Snippet controls. noindex, nosnippet, max-snippet:0 and data-nosnippet keep content out of AI answers on Google and Bing. Templates often apply them by accident.
• Bing Webmaster Tools. Microsoft Copilot grounds its answers on the Bing index, so Bing health is no longer optional.

Per-bot access

OpenAI documents GPTBot as its training crawler and OAI-SearchBot as the crawler for ChatGPT search, and says each setting is independent. A blanket AI block removes you from ChatGPT search. You can opt out of training and stay in search:

User-agent: GPTBot
Disallow: /

User-agent: OAI-SearchBot
Allow: /

Firewall and CDN testing

OpenAI's help centre says eligibility needs both: robots.txt must allow OAI-SearchBot, and your host or CDN must allow its published IP addresses. A firewall can return 403 to a crawler robots.txt allows, and no dashboard shows it. Cloudflare began blocking Training and Agent crawlers by default on ad-displaying pages for new domains and free-tier customers from 15 September 2026, so sites can lose access without editing anything. The AI Crawler Checker requests a page as each AI crawler and shows which ones get turned away.

Fact checks in AI answers

AI platforms get specific facts wrong about specific brands: prices, product names, ownership. The fix starts on your own site. Remove contradictions between pages and state key facts in plain HTML a crawler can fetch. Measure each engine separately. In Otterly's 2026 experiment, Microsoft Copilot produced 2,718 of the 3,934 tracked citations for one site while ChatGPT and Google's AI surfaces barely cited it.

Primary sources contradict these, so treat them as signs of outdated advice:

• "Add schema or FAQ schema to get cited." Google says structured data isn't required for generative AI search. Ahrefs' matched test found AI Overview citations fell 4.6% after pages added schema. Existing markup does no harm. It just isn't an AI lever.
• "Publish an llms.txt." Google says Search ignores it, and no other engine documents reading it.
• "Improve your E-E-A-T score." Google says E-E-A-T isn't a specific ranking factor. There is no score.
• "Build domain authority." It's a Moz metric. No engine computes it.
• "Allow GPTBot so ChatGPT finds you." GPTBot governs training only.
• An AI rank or position in Google. Google reports AI impressions only, with no position data.
• Guaranteed citations. See why citation guarantees don't hold up.

1. Do you check robots.txt per bot, separating training crawlers from search crawlers?
2. Have you requested our pages as OAI-SearchBot, PerplexityBot and Claude-SearchBot, and what status codes came back?
3. Is our pricing in the raw HTML before JavaScript runs?
4. Are any templates applying nosnippet or data-nosnippet by mistake?
5. Are we verified and healthy in Bing Webmaster Tools?
6. Which facts do AI engines get wrong about us, and on which engine?
7. What will you not promise, and why?

Good answers name a test and a result. For a longer list, see how to vet an AI SEO agency.

Send the questions above before your next renewal. If the agency covers the three gaps, ask it to add them to scope and report results per engine, as in this AI visibility client reporting playbook. If it can't, compare a specialist add-on against doing the work in-house, and check pricing against what a fair AI SEO package covers.

For a baseline first, the free checker on this site shows whether AI engines name you for up to three buyer questions. It is a snapshot, not tracking.

Be clear about limits. An agency can remove blocks. It cannot control what a model learned in training, which pages an engine retrieves, or what it chooses to cite.

Should I pay my SEO agency extra for AI search work? Only for the new tasks: per-bot access rules, firewall and CDN testing, and per-engine fact checks. Indexing, rendering, snippet controls and Bing hygiene are already part of competent SEO, so a large AI surcharge for those deserves questions. Ask what each added line item tests and what result it reports.

Should we remove schema markup we already have? No. Google says existing structured data does no harm. The point is narrower: schema is not a lever for getting cited in AI answers, and Ahrefs' matched test found AI Overview citations fell slightly after pages added it. Keep what supports normal rich results and stop paying for new schema sold as an AI fix.

Can my agency report my ranking in Google AI Overviews? Not with real position data. As of October 2026, Google's Search Console report for its AI features shows impressions only, with no clicks, click-through rate or position. Any AI rank an agency reports comes from its own tool and has no Google data to check it against. Ask how the number is produced.

Does blocking GPTBot remove us from ChatGPT search? No. OpenAI documents GPTBot as a training crawler and OAI-SearchBot as the crawler for ChatGPT search, and says each setting is independent. You can block GPTBot to opt out of training while allowing OAI-SearchBot. Your firewall must also let OAI-SearchBot through, not just robots.txt.