Nearly two‑thirds of Americans view AI development as a threat, new poll shows

Almost two in three Americans now say the development of artificial intelligence is a bad thing, according to a new poll released this week. The finding, published by The Hill, marks a sharp rise in public unease and comes as companies pour record spending into AI projects and experts warn of mounting societal risks.

Survey findings that capture public sentiment

The Hill’s survey of U.S. adults asked respondents whether they believed AI development was overall good or bad. “Almost two in three say AI development a bad thing,” the headline read, indicating a clear majority of respondents view the technology with suspicion. The poll also revealed that many respondents trust AI less than other emerging technologies and express worries about job displacement, privacy erosion, and the potential for AI‑generated misinformation.

While the exact numeric breakdown was not disclosed beyond the “almost two in three” phrasing, the result aligns with a broader trend highlighted by Pew Research Center’s 2026 study of Americans’ attitudes toward AI. Pew reported that a sizable share of the population believes AI will have a “mostly negative” impact on society, and that confidence in AI’s benefits is outweighed by concerns over its risks Pew Research Center.

Industry’s parallel trajectory

At the same time, McKinsey & Company’s “State of AI in 2026” report paints a picture of accelerating corporate investment. The consulting firm notes that firms across sectors are moving from pilot projects to broader deployments, seeking measurable returns on AI spend. Yet McKinsey cautions that “most organizations are still in the early stages of realizing AI‑driven ROI,” and that the gap between ambition and actual financial outcomes remains wide McKinsey & Company.

Similarly, an Ipsos data set on artificial intelligence highlights that while businesses are eager to adopt AI tools, consumer confidence lags behind. The Ipsos tables show a divergence: enterprises report optimism about efficiency gains, whereas the public remains wary of unintended consequences Ipsos.

Why it matters

The clash between public apprehension and corporate enthusiasm creates a policy dilemma. If a majority of citizens view AI development as a negative force, democratic pressure may push regulators to impose stricter oversight, potentially slowing the pace of innovation. Conversely, firms that continue to invest heavily could outpace regulatory frameworks, leading to a “wild west” environment where ethical lapses become more likely.

Experts surveyed by MIT Sloan’s “Most urgent AI risks” study underscore the stakes. Among 272 specialists, the top concerns included “bias and discrimination,” “misinformation and deepfakes,” and “loss of human control over autonomous systems.” The consensus was that without coordinated governance, these risks could translate into real‑world harms that erode public trust MIT Sloan.

Legal professionals echo the warning. A Thomson Reuters piece on AI’s role in law in 2026 reports that lawyers are grappling with questions of liability, evidentiary standards, and the ethical use of AI in courtrooms. Practitioners argue that the law is playing catch‑up, and that the current regulatory vacuum could exacerbate the very concerns expressed by the public Thomson Reuters.

Comparing the data: consensus and gaps

All six sources converge on three core observations:

1. Public sentiment is increasingly skeptical of AI’s net impact.
2. Corporate investment in AI is rising, but measurable ROI remains limited for many firms.
3. Experts identify a suite of high‑stakes risks that have yet to be fully addressed by policy.

Where the sources diverge is in the magnitude of optimism among businesses. McKinsey emphasizes “rapid scaling” and “significant upside,” whereas Ipsos points to a more cautious outlook, noting that many firms still treat AI as a “support function” rather than a core driver of growth. The Hill poll, focused on public opinion, does not provide industry‑specific data, leaving a gap in understanding how consumer concerns translate into purchasing or adoption behavior.

Another point of contrast lies in the timeline for risk mitigation. MIT Sloan’s expert panel calls for immediate action on bias and misinformation, while the Pew study suggests that public attitudes may shift slowly as familiarity grows. The legal community, meanwhile, stresses the need for “interim frameworks” to guide AI use in litigation, indicating a shorter‑term urgency.

Looking ahead

Several near‑term milestones will likely shape the trajectory of AI development and public perception:

• By the end of 2026, major tech firms have pledged to publish “model cards” detailing the capabilities and limitations of their most advanced generative AI systems, a move intended to increase transparency.
• Congress is expected to hold its first comprehensive hearing on AI risk governance in early 2027, with testimony from industry leaders, ethicists, and consumer‑advocacy groups.
• McKinsey projects that AI‑related spending will surpass $200 billion globally by 2027, though it cautions that “only a minority of that spend will be tied to proven ROI” at present McKinsey & Company.
• Legal firms anticipate the rollout of the “AI‑in‑Court” guidelines by the American Bar Association in mid‑2027, aiming to standardize the admissibility of AI‑generated evidence.

These developments will test whether the growing public unease can be addressed through clearer standards, or whether the momentum of corporate investment will outstrip societal safeguards. The next year will be pivotal in determining whether AI’s promise is realized as a driver of economic growth or as a source of new social challenges.