The answer looked plausible
Weaver searched Google for “snipe hunting season in Alaska.” The AI Overview at the top of the results said the season began on September 1.
That date applies in some areas of Alaska. But not Kodiak. After the October 8 opening of the general waterfowl season made her question whether her September hunt had been legal, Weaver checked the regulations and contacted an Alaska wildlife officer.
The officer’s written statement said his own search for “snipe hunting season” produced similar results. He also said a closer check of the regulations would have shown Weaver that snipe hunting was permitted from October 8 through January 22.
Snipe are small, stocky shorebirds with long bills, common across North America. Their population is stable. By late summer, they have finished breeding and begun migrating south.
A search answer is not a regulation
Weaver told Alaska Beacon she was deeply ashamed and struggling with what had happened. She said that after returning home and reviewing the rules, she reported herself immediately.
Google did not respond to Alaska Beacon’s request for comment. The publication also cited an analysis arguing that Google AI Overviews may spread inaccurate information at a scale with few precedents in human history.
I think the revealing detail is that the wrong date was not obviously absurd: it was valid somewhere in Alaska. That makes this less a case of a system inventing a rule than of a concise answer flattening a rule that changes by location. The missing test is whether the summary makes that boundary clear before a reader acts on it.
The cost of a wrong answer here was not just a bad search result. It was a person trying to work out whether she had broken the law, while the system’s confident summary left the crucial qualifier out of view.
Daily AI news
Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.
Only what matters — every day
Follow on X