The publishing of an old article about a 2002 bankruptcy-court filing by United Airlines (UAL) on Google news on Monday has highlighted the downside of using automated and algorithmic trading applications to execute orders.

It still remains unclear how the six-year old report of UAL's 2002 bankruptcy-court filing appeared on the ‘most viewed' section of Sun-Sentinel's Web site on Monday. But the item was picked up by the Google news crawler and appeared on the Google news site as a fresh report that day.

It is thought that the out-dated story was then picked up by other news aggregators and was seen by a Florida-based investment firm which posted a summary on the Bloomberg financial information service. The posting on Bloomberg prompted automated trading programs to dump UAL stock and resulted in the UAL share price crashing from $12 to $3 before Nasdaq halted trading.

The incident has raised new concerns about the automated systems that are used to both gather news and to buy and sell stocks based on headlines and earnings data. According to a Wall Street Journal report this kind of algorithmic trading was responsible for around a quarter of New York Stock Exchange trades in the last week of August.

But this incident shows clearly that machines can't always determine if a story is worth acting on and sometimes human intervention is vital.

Bernie McSherry, a senior vice president with New York institutional brokerage Cuttone & Co, told the WSJ that a trader can "pull back" before proceeding, but some of the less sophisticated automated trading systems can't do that.

Automated trading bites some UAL investors - Reuters