As of November 30, 2025, we are only ~14 months into a 2–3 year prediction window that runs roughly from September 2026 to September 2027 for the full outcome. It is therefore too early to decisively judge whether “massive disruption” has occurred or will occur in that time frame.

What we see so far (supports possible future disruption):

• Multiple surveys indicate that a majority of contact centers already use some form of AI (chatbots, assistants, analytics), with estimates around 50–65% adoption, and large shares planning further investment. (emerge.haus)
• Industry reports project rapid growth in call-center AI markets through the late 2020s, with strong economic incentives for automation and workforce reduction. (businessresearchinsights.com)
• Some forward‑looking analyses specifically forecast that by 2027 leading call centers could have 50–60% smaller human workforces due to AI, implying substantial disruption if these projections materialize. (emerge.haus)

What we don’t yet see (why it can’t be called “right” today):

• Current credible estimates still show most customer interactions handled by humans, with AI automating a minority of volume. Gartner, for example, projects only about 10% of agent interactions fully automated by 2026, which is significant but not obviously a “massive” displacement of the industry as a whole. (sci-tech-today.com)
• A 2025 Cavell study projects that global contact-center agent roles will increase from 15.3M (2025) to 16.8M (2029), concluding that AI will slow agent growth rather than cause net job collapse in the near term. (cxtoday.com)
• A Goldman Sachs–linked survey finds that, as of late 2025, only about 10% of firms have actually cut jobs due to AI, and many executives still consider AI too early for broad deployment, even though they expect more impact over the next few years—especially in customer support. (businessinsider.com)

Why the verdict is “inconclusive (too early)”

• Sacks’s prediction is about what will happen within 2–3 years, not what must be true just one year in. For it to be clearly wrong, we would need to reach at least the 3‑year mark (September 2027) without seeing industry‑wide, large‑scale displacement or transformation.
• For it to be clearly right already, we would need strong evidence that by late 2025 a “massive disruption” had already occurred—e.g., large, measurable drops in global call‑center employment or a clearly dominant share of interactions being AI‑handled. Current data instead shows rapid adoption and meaningful efficiency gains, but still incremental displacement and even net growth in agent headcount.

Given the remaining time in the forecast window and the mixed but rapidly evolving data, the only defensible status as of November 30, 2025 is “inconclusive (too early)”, not definitively right or wrong.