Australian banking giant, Commonwealth Bank of Australia (CBA), was forced to backtrack on a layoff last month after a tribunal hearing, in a serious PR blunder that has seen 45 employees reinstated. The layoffs were first announced after the bank said artificial intelligence rendered the employees’ positions redundant.

CBA later admitted the staff didn’t have to go after a union tribunal challenged that assertion. It came after the Finance Sector Union (FSU) said CBA “misled staff and the community” by claiming the bank’s new AI-based “voice bot” was responsible for the lost jobs, when in reality the reduction in call volume was not true.

It added that while the bank said its introduction of a new “voice bot” reduced call traffic by up to 2,000 per week, the reality was that staff have seen call volumes increasing while many of those laid off had been at the company for three decades.

Finance Sector Union, which had brought the case against CBA, said that “following a hearing last week, CBA has accepted that it did not consult appropriately before telling our members their jobs were redundant”, indicating that management had skipped steps in its attempt to reduce labor costs.

The bank has since admitted to the tribunal that call volumes increased before the staff layoffs and during the period of the redundancy process, in sharp contrast to the image it was trying to project. It has been revealed that the bank had to ask managers in the contact center to help with the increasing workload at the time. The spike in call volume at the time of layoffs has been attributed to a change in the government’s policy on transfer payments. It was revealed in a submission by CBA to the tribunal that “these changes caused an unanticipated and significant increase in calls lasting several months, with a corresponding impact on staffing requirements, including overtime being offered”.

It should be noted that at the time the layoffs were announced, the bank claimed that it did not expect this to have a long-term effect on the number of workers at the bank, as the effects of the “voice bot” would impact the full-time equivalent staff count, but not the number of permanent full-time positions. It is with this logic that it is said that the redundancies, when combined with “normal turnover”, would have no impact on the future number of positions.

“The damage has already been done; some members haven’t worked and have been unable to pay their bills,” the union said, in a statement to the press. It was later admitted that the redundancies were made without genuine consultation. It added “this is despite management knowing the significant risk of an increase in call volumes, either as part of the testing of the voice bot, or due to other factors”, Bloomberg reported.

The admission by the CBA follows its pledge in a recent report to take up investment in open source technology like ChatGPT. In a recent partnership with Microsoft, CBA agreed to accelerate AI investment into both new and existing products.

This development came just last week as CBA announced the launch of a new partnership with OpenAI AI which will see the two companies work together to develop generative AI solutions for banking in areas such as fraud detection, anti-money laundering, and personalized customer experiences.

“We’re not using AI to replace people, we’re investing in our people and technology to drive a better, more personalized customer experience and protect our customers from scams,” the bank said of the partnership. However, trust has been badly broken after what has transpired, the union added.

Executives are now in panic mode and are desperately seeking to remove what they now think may be vulnerabilities in their plans to automate staff roles. Bank of America and Wells Fargo are also partnering with OpenAI, while JPMorgan is also said to be expanding its ChatGPT partnership with Microsoft.

The trend has been for the financial sector as a whole to shrink staff numbers as some 200,000 jobs could be eliminated in three to five years as artificial intelligence is expected to disrupt “back office, middle office and operations functions”, Bloomberg Intelligence estimates. AI could save billions, but that may not include the cost of doing PR damage control to ensure staff still trust bank management.

The new partnerships will start with smaller-scale proof of concepts “focused on delivering responsible generative AI applications to our business”, CBA’s chief information and digital officer, Ian Sloan, said in the statement. However, whether or not the bank will still stick to its original plans with staff on edge will be determined in the coming weeks and months, as staff have already signalled their mistrust in management.

The FSU claimed that at the time of layoffs, the bank had not explained the basis on which roles had been assessed to be redundant, which raised an interesting question on whether the layoffs were at all needed. It is alleged that management may have decided to move some of those jobs offshore to India instead and therefore hired staff in India to fill those roles. The move made the new chatbot AI sound like a convenient excuse.

Sources close to the CBA told Bloomberg they “won’t be losing a minute’s sleep over this. The plan has been and still is to go ahead and outsource some of the roles. The Union just got in the way”, in a possible hint of things to come.

“The U-turn on 45 jobs is a massive win for our members, but it’s hard to ignore the obvious damage that’s already been done. Many members were faced with financial hardship; some didn’t work for weeks, while those on board were left in limbo,” union spokesperson told the outlet. In a concession to stave off the union protest, CBA said that it has now offered “voluntary redundancy, redeployment, or retention” to the staff. “We have apologized to the employees concerned and acknowledge we should have been more thorough in our assessment of the roles required,” a spokesperson said.