Hundreds of staff at TikTok’s Irish arm were given competency tests in late 2023 without knowing that the results would go towards deciding whether they would keep their jobs in a mass redundancy drive the following spring, a tribunal has heard.
The Workplace Relations Commission (WRC) heard evidence on the tests this week as part of a challenge by a former content moderation team leader who lost his job last year after missing the cut to keep his job by a fraction of a point.
Over 1,900 workers worldwide who were put at risk of redundancy in early 2024 were quizzed on their knowledge of TikTok’s content moderation policies – some 950 of them in Ireland, the Workplace Relations Commission was told.
A senior manager at the social media firm said bosses had taken a “quite deliberate” decision to keep the workers and their line managers in the dark about the true reasons for the test.
Trump’s on-off tariff announcements now the main threat to global growth
‘The mask is off in tech. You’re getting fired if you speak out’
Child Benefit: Will payments rise for lower income households and what will this mean for wealthier parents?
Chambers to focus on costs over design for big projects and how Trump could hit the economy
The tribunal was hearing evidence on a complaint against TikTok Technologies Ltd under the Unfair Dismissals Act 1977 by Mohur Saleh, who lost his job in April 2024.
Mr Saleh was one of 564 staff made redundant worldwide at that time by TikTok, 289 of them in Dublin, the WRC was told.
However, Mr Saleh has argued he was unfairly selected for redundancy on the basis that the scoring system in his talent pool was partly based on the results of the November 2023 test – which he had not passed – as well as on performance ratings, which he had disputed.
The WRC heard Mr Saleh scored 32.25 points as part of a redundancy selection process which began in February 24. That was 0.75 short of the cut-off point at 33 points.
IATA Director General Willie Walsh on airline profits, air fares and why the Dublin Airport passenger cap makes Ireland a laughing stock
Mr Saleh had scored 25 points for his clean disciplinary record, with his performance ratings for 2022 and the first half of 2023 accounting for the balance of his score, the WRC heard.
However, Mr Saleh did not make the minimum grade in the policy knowledge test and was awarded 0 out of 10 available to him, the tribunal heard.
Kevin Purcell, who was head of training and quality for TikTok in Europe, the Middle East and Asia (EMEA) at the time of Mr Saleh’s redundancy, said that senior management “formed a view that we needed to restructure the entire organisation globally”.
That meant it required only 39 out of the 63 moderator team leads on staff and was aiming to shed 24 positions, the tribunal heard.
Mr Purcell said he had no knowledge of which staff “made the cut” until the points were calculated and ranked.
He said there was “clear water” between Mr Saleh’s position and the cut-off point as a number of other employees scored higher than the complainant but failed to secure one of the available jobs.
Over 1900 workers in the areas later affected by the redundancies were directed to do the policy knowledge tests in November 2023, the tribunal heard.
The workers were asked to review 100 pieces of content which had already gone through TikTok’s content moderation process, with their answers being checked against the official moderation outcome, the tribunal heard.
Mr Purcell confirmed when questioned by the adjudicator, Monica Brennan, that the tests were done “with the restructuring in mind” and that staff did not know the purpose of the policy tests at the time.
“That was quite deliberate, because it was a blind test. I was concerned that if people knew it was being done in contemplation of restructuring, it might lead to people cheating,” he said.
Mr Purcell said the pass grade in the test was 51 out of 100 and that Mr Saleh had failed to pass this threshold.
Mr Saleh’s evidence was that 35 of the 100 posts did not load when he attempted the test.
He also took issue with the fact that the test was based on TikTok’s rules for the English-language market, which were different from the rules for the Middle East and North Africa where he had primarily worked.
Mr Saleh said his separate performance ratings had been affected by “misconduct and harassment, including serious harassment, involving colleagues”, with a knock-on effect on his performance, and about which he had complained.
He said positive feedback about projects he had undertaken ought to have been taken into account, in 2022 and 2023, but wasn’t. He said he had looked for a review of his 2022 appraisal without success. Mr Purcell told the WRC in his evidence that he had been involved in “calibrating” the performance review process in TikTok ahead of the mid-2023 round.
He said that before that point, the reviews were “very skewed towards positive across the board”. He said he felt Mr Saleh’s rating was “appropriate” in 2023 and that the “recalibration” of performance reviews “wasn’t directed to the complainant or any individual”.
The WRC heard that during the redundancy consultation period, Mr Saleh applied for four internal roles at the team leader level or higher, but did not apply for any open vacancies at the lower rank of individual contributor.
Having failed to secure an alternative position, he was made redundant, the tribunal heard.
In a legal submission, TikTok’s barrister, Niamh McGowan BL, appearing instructed by A & L Goodbody, said Mr Saleh lost his job as part of a “mass redundancy” which had been carried out in consultation with elected employee representatives.
“He’s the only employee that has challenged the fairness of the process and the selection criteria. What we say is that the dismissal arose entirely from redundancy,” Ms McGowan said.
“There was an open, meaningful and fair collective consultation process which applied equally to every other person put at risk,” she added.
The adjudicator, Ms Brennan, is now considering her decision on the case, which will be published by the WRC in due course.
