AI BriefWire / Briefing
“We are not querying like humans anymore”: OpenSearch’s Bianca Lewis on why AI agent mistakes look like successes
On 2026-11-20, The New Stack AI reported that OpenSearch’s Bianca Lewis discussed how AI agent mistakes can look like successes. The supplied excerpt says an AI agent may call the wrong API and receive a seemingly valid result.

Full analysis
What happened, why it matters, the business impact, and what operators should watch next.
What happened
On 2026-11-20, The New Stack AI reported that OpenSearch’s Bianca Lewis discussed how AI agent mistakes can look like successes. The supplied excerpt says an AI agent may call the wrong API and receive a seemingly valid result.
Why it matters
A seemingly valid result may make an incorrect AI-agent action difficult to recognize, according to the supplied report.
Business impact
Organizations using AI agents may need to consider how to identify incorrect API calls that produce apparently valid results. The supplied facts do not quantify operational, financial, or customer impact.
Who is affected
Teams tracking AI Agents, Agents, product strategy, operations, and market positioning.
Operator take
Consider evaluating observability and validation practices for AI-agent API calls, subject to further evidence and organizational requirements.
What to watch next
Consider evaluating observability and validation practices for AI-agent API calls, subject to further evidence and organizational requirements.