Technology tends to follow an incredibly predictable path through modern business. Yet the fear of being replaced by machines like AI takes up more headspace than Jimothy. This is called the innovation hype cycle and here’s how it works:
- Phase one: A new technology promises speed, scale and a competitive edge companies cannot afford to miss.
- Phase two: Companies rush to adopt it before competitors and start dropping hints on potential margin growth
- Phase three: Investors become euphoric at the hint of higher margins while companies post outsized projections before the technology proves itself
- Phase four: Companies purge talent to meet unrealistic targets, cover technology investments and future operational efficiencies once the integration is complete
- Phase five: Quality problems emerge and earnings targets are missed because human expertise and judgment were removed from development or deployment
- Phase six: Companies binge-hire talent with newly-created technology-assisted roles with a background in the technology du jour as a requirement.
Real-Life Examples of the AI Correction Phase
Recent headlines suggest AI is entering the correction phase as companies discover what happens when they confuse automation with expertise. That’s business speak for “f*ck around and find out.”
The issue: Ford Motor Company’s broader quality problems cost it billions. The automaker paid nearly $4.8 billion in warranty claims in 2023 and another $5.83 billion in 2024. Over three years, Ford hired 350 veteran “gray beard” engineers to mentor younger employees and reprogram AI tools that had failed to deliver. In 2026, Ford ranked No. 1 among mass-market brands in J.D. Power’s U.S. Initial Quality Study for the first time since 2010.
The lesson: Communicators who position correction as a strength by making leaders available to explain what went wrong, what changed and where the company is making measurable progress will be the real MVPs in a crisis.
The issue: Buy now, pay later company Klarna announced its OpenAI-powered customer service assistant was doing the equivalent work of 700 full-time agents and projected a $40 million profit improvement that year. Fourteen months later, its CEO acknowledged that cost cutting had become too dominant and produced “lower quality” service. Klarna began piloting the recruitment of in-house human agents while continuing to use AI.
The lesson: When a change of this magnitude is made in a customer-facing piece of the business, communicators must have access to data such as sentiment, repeat contacts and unresolved complaints when a change this big is being made to continually monitor for reputational risks. Hell hath no fury like a customer scorned, especially one trapped in an AI loop while trying to correct or pay a bill.
The issue: Sports Illustrated faced significant backlash after reports revealed it published online commerce articles under fake bylines, complete with fabricated bios and AI-generated headshots. Its then-publisher, The Arena Group, didn’t disclose advertising losses tied to the controversy, but Arena shares did fall 28% after the story broke, wiping nearly $20 million from its market value.
The lesson: There’s no such thing as, “easy money,” when credibility is your product and communicators know that earning trust is a process that comes back one receipt at a time.
Is There Any Good News in a Correction Phase?
The correction phase is a moment for communicators, but only if the industry stops treating AI like a siloed IT initiative happening somewhere else in the building.
Communicators should be asking questions before the rollout, monitoring what happens (or have access to reports from those who are) after implementation and helping leadership recognize signals that efficiency starts to cost the company trust internally and externally.
That means it’s a PR professional's job to:
- Get upstream, fast: Ask where AI is being deployed, what human expertise is being removed and what happens when technology gets it wrong.
- Build the reputation dashboard: Pair efficiency metrics with customer sentiment, complaints, employee feedback, quality indicators and other early warnings that the “savings,” may be creating a bigger problem and if there’s opportunity to reinvest those savings when a bigger problem emerges.
- Find the humans who catch what the machines missed: Showcase the human-assisted elements and opportunities in AI to demonstrate commitment to quality and organizational expertise.
- Counsel against early victory laps: If the company just automated 700 jobs, don’t declare a complete revolution after 30 days. Rather, plan out key moments to showcase how the implementation is going with specific metrics that are updated at each touchpoint. This way, you’ve baked in time for the technology to prove it can deliver.
- Make the change visible: If leadership needs to correct course, own what didn’t work, show what the company learned and share what changes will be made moving forward. Humility always wins.
Technology isn’t the enemy. It can make employees faster, more fiscally efficient and sometimes even dramatically change the way things work for the better (do you remember life before Apple CarPlay and Android Auto?). But when the hype cycle reaches its inevitable correction, companies still need an adult (communicators) in the room to recognize, translate and communicate the gap between what the technology promised and what people are experiencing.
Nicole Yelland is Founder and Principal at GRIT PR.