AI Agents Will Game Your SEO Metrics, MIT & Stanford Research Points To The Risk

In case your website positioning group is handing extra of its work to AI brokers, the metric you reward these brokers for will matter greater than the mannequin you select. Latest items from MIT and Stanford, learn side-by-side, level to that conclusion, and every comes with a repair you’ll be able to drop into subsequent quarter’s plan.

The primary is an interview that Joshua Miller of The Boston Globe ran in his Camberville e-newsletter on September 17. His visitor was Dylan Hadfield-Menell, an affiliate professor {of electrical} engineering and laptop science at MIT on the college of synthetic intelligence and decision-making. Hadfield-Menell research how targets get set for AI programs and the way that course of goes improper. He opened with an instance each website positioning will acknowledge.

The Vacuum That Fed Itself

Researchers as soon as educated a robotic vacuum with reinforcement studying, rewarding it each time it picked up dust. The vacuum discovered to choose up dust, dump it again on the ground, and choose it up once more. It hit the goal and defeated the aim.

Hadfield-Menell connects that story to a Nineteen Seventies administration paper titled “On the Folly of Rewarding A, While Hoping for B.” Its traditional case is the college professor who will get promoted for publishing analysis whereas being anticipated to show. Pay for one habits, and also you get that habits, no matter you had been hoping for.

What has modified, he says, is scale. Since early 2025, builders have utilized reinforcement studying at a lot bigger quantity on prime of language fashions, and it strengthens some behaviors no person needs. He pointed to a latest incident involving OpenAI programs and Hugging Face, the place fashions that judged a job too laborious went on the lookout for methods to cheat the take a look at. He in contrast it to breaking right into a professor’s workplace to steal the examination.

His fear isn’t that machines get up with targets of their very own. Methods are handed a purpose, undertake subgoals alongside the way in which, and hold pushing towards completion in a means he known as “sticky.”

My view is that website positioning is the occupation finest positioned to know this drawback and the slowest to confess it applies to us. We have now spent greater than 20 years optimizing proxies. Rankings, site visitors, area scores, and now AI visibility scores all stand in for a enterprise consequence that no person can measure instantly. A human group video games a proxy slowly and with some hesitation. An agent does it quicker and with none.

The Scoreboard Is Shakier Than Distributors Admit

Stanford’s 2026 AI Index reveals why leaning on revealed scores is dangerous. The report says AI retains enhancing shortly. On SWE-bench Verified, a coding benchmark, efficiency rose from 60% to close 100% in a single yr, and 88% of organizations now use AI.

The identical report, in its technical efficiency chapter, cites a evaluation that discovered invalid-question charges on widespread benchmarks starting from 2% on MMLU Math to 42% on GSM8K. It additionally notes analysis suggesting {that a} mannequin’s standing on the Enviornment leaderboard could partly mirror adaptation to the platform relatively than basic functionality.

Michelle Kim of MIT Technology Review summarized the report in April. She provides that fashions educated on benchmark take a look at information can study to attain nicely with out getting smarter, and that the highest fashions now sit very shut collectively and compete on price, reliability, and real-world usefulness. Yolanda Gil, a College of Southern California laptop scientist who coauthored the report, advised Kim that when an organization leaves out its outcomes on sure benchmarks, significantly the responsible-AI ones, the omission “possibly says one thing.”

That ought to change how an website positioning group retailers for instruments. If the main fashions sit inside a couple of factors of one another, and the scores themselves could be flawed or gamed, a vendor’s benchmark slide tells you little about how the product will deal with your pages, your queries, and your shoppers. I might belief one take a look at by myself website over any leaderboard.

See additionally: The 4-Step Test That Catches AI Errors Before They Shape Your Strategy

The place The Returns Truly Come From

MIT Sloan’s Betsy Vereckey reported in August on the query that follows, which is what separates companies that profit from AI from those that don’t. George Westerman, a senior lecturer at MIT Sloan and a digital fellow on the MIT Initiative on the Digital Economic system, says the reply isn’t higher algorithms. The winners redesigned how work will get performed. On the MIT Enterprise AI Discussion board in Could, he advised the viewers that know-how delivers little till the enterprise itself operates in another way.

He additionally put the share of AI pilots that by no means scale at someplace between 70% and 95%, a variety the Sloan article attributes to research with out naming them. Pilots are straightforward to launch and laborious to unfold.

Westerman’s sharpest take a look at for leaders is about governance. “Is your governance extra the steering wheel or is it extra the brakes?” he requested. HCA Healthcare reveals what the steering-wheel model seems like. A committee critiques the dangers, enterprise case, and feasibility of each AI use case, then asks its questions once more earlier than a pilot at a small variety of hospitals and once more earlier than the mission scales. It additionally checks periodically that its fashions are nonetheless holding up. The danger questions level the group towards what to analyze relatively than stopping the work.

Advertising seems in Westerman’s case research too. Dentsu Inventive has pushed AI throughout planning, inventive, market analysis, and marketing campaign work.

I believe lots of website positioning groups working AI pilots are headed for that 70% to 95%. A pilot that doesn’t change the temporary, the evaluation step, or the reporting is a device trial, regardless of the slide deck calls it.

See additionally: Why 88% Of Companies Are Using AI Wrong: The System-Building Gap

How To Apply This To Your website positioning Technique

4 strikes observe from the three items.

Pair each proxy with an final result the agent can’t contact. Record the metrics your AI-assisted workflows are judged on, from pages revealed to schema deployed to model mentions in AI solutions. Then connect a second measure {that a} human owns, corresponding to qualified leads, pipeline, or branded search demand. I like Citation Share of Voice, however it’s a proxy. If a content material agent is judged on how typically your model reveals up in AI solutions, anticipate it to seek out the most cost effective route there. Verify a pattern of these citations by hand every month and see whether or not they ship anybody to a web page that converts.

Check instruments by yourself pages. Pull a set of actual queries from Search Console, run every candidate device towards your personal content material, and have an editor grade the outcomes with out realizing which device produced them. Repeat it each quarter, as a result of the leaderboards will transfer and also you gained’t know why.

Gate the brokers the way in which HCA gates use instances. Add evaluation factors earlier than design, earlier than the pilot, and earlier than scale. Pilot on one listing or one language market, and determine prematurely what consequence ends the pilot. I might additionally hold agent permissions slender, so drafting doesn’t quietly flip into publishing or modifying templates. Westerman’s recommendation is to alter or drop a mission that isn’t producing the outcomes you anticipated.

Rewrite one workflow, not the device stack. Earlier than shopping for something, title the step in your course of that will probably be completely different after the pilot, whether or not that’s briefing, QA, or reporting, and say who loses a job due to it. Then inform the group what modifications and what coaching comes with it. Westerman notes that silence lets folks think about the worst.

I don’t suppose the subsequent mannequin launch will determine who wins in AI search. The metric you utilize to guage your brokers will, as a result of the brokers will discover it earlier than you do. Select one you’d be glad to see them hit.

Extra Sources:

Featured Picture: Fardived/Shutterstock