Andus Labs

Andus Labs

By Andus Labs
  • The Index ranks the 25 most urgent patterns stalling enterprise AI. Nearly half trace back to a leadership decision, or the lack of one.
  • The top-ranked pattern, Empty Chairs, describes companies giving AI a say in business decisions like hiring, pricing and risk without naming who is accountable.

NEW YORK, Sept. 16, 2026 /PRNewswire/ -- Andus Labs today published Release 02 of its Ground Truth Index, ranking the top 25 barriers standing between enterprise AI investment and return. Nearly half of this Index's patterns cluster around leadership decisions, including seven of the top ten. Empty Chairs ranks first, with AI systems making decisions no one has clear standing to review or authority to change, and that's creating new risk. Learning While Drowning ranks second, where employees are told to learn and given no hours to do it.

Andus Labs

Twenty of the 25 are newly elevated from Andus Labs' library of 300 documented patterns; five carried over from the inaugural Index. Release 01 acknowledged that the models worked and put the challenges with people who lacked the trust, skill or motivation to use them. The five that carried over — Shadow Agency, Unchecked Output, Pilot Graveyard, Reality Resistance and Skill Atrophy — describe what goes wrong when usage grows without governance from the top. Four now rank in this release's top ten.

Six patterns also describe employees who appear to be resisting the technology but are responding rationally to conditions their company set. In those cases, the fix also belongs with leadership.

"The models work. What's breaking is everything around them," said Chris Perry, Founder and CEO of Andus Labs. "Companies are still asking whether AI is capable enough, when the harder questions are whether anyone owns the decisions these systems make, whether employees have the time and incentive to work differently, and whether the organization is built to absorb what the technology can now do. Ground Truth Index names these problems precisely enough that someone can own them."

The five highest-ranked patterns:

Empty Chairs. Companies give AI systems greater decision-making power without naming a specific person responsible for overseeing, and sometimes overturning, those decisions.

Learning While Drowning. Organizations add mandatory AI training to already-full workloads without making room for employees to learn new ways of working.

Optimization Temptation. AI makes the savings from cutting headcount easy to quantify. The gain from moving those same employees into higher-value work is harder to measure, as is the cost of the institutional knowledge that leaves with them.

Unfunded Ambition. Companies declare aggressive AI goals without putting sufficient budget, authority or changes to how work gets done behind them.

Jargon Over Capability. Organizations adopt the language of AI faster than they build the systems and workflows that language describes.

"Every company has an AI strategy. Almost none of them wrote it deliberately," said Mike Connery, Partner, Intelligence. "Across all 25 patterns, five authors carry more influence than leadership in setting it: the market, competitors, vendors, the org chart, and the workforce."

Over the past 18 months, Andus Labs has documented these patterns through client engagements, prospect conversations and executive forums with leaders at 595 organizations across financial services, healthcare, manufacturing, technology, energy, consumer goods, and beyond. Business leaders describe what's happening in their organizations, and Andus Labs researchers name the patterns underneath. Patterns are added, merged and retired every quarter.

The Index is human at the start, human at the end, and AI in the middle for scale. For each release, five independent analyst agents score every pattern in the library against five dimensions: gravity, velocity, proximity, progression, and scale probability. Agents cannot see each other's results, and each score is the median of five reads. The top 50 then undergo researcher review and adjudication before the final 25 become the Index.

Each pattern includes a diagnostic question leaders can use to test whether it's present in their organization.

The latest Ground Truth Index is available at [https://humanreadiness.org/ground-truth-index]. Release 03 publishes in November.

About Andus Labs

Andus Labs builds the Human OS for AI, the operating layer that orchestrates work across people and agents. The work spans operational redesign, decision simulations, and agentic products that change how enterprises run, with human judgment and accountability intact. Andus Labs delivers services-as-software to Fortune 500 clients and global IT leaders. The Ground Truth Index is the company's public diagnostic, drawn from a library of 300 barriers to enterprise AI observed in the field.

Media Contact

andus@haymaker.co

Cision View original content to download multimedia:https://www.prnewswire.com/news-releases/andus-labs-latest-ground-truth-index-finds-top-5-barriers-to-enterprise-ai-start-with-leadership-not-the-workforce-302879959.html

SOURCE Andus Labs