Before Information Technology (IT) became a mature multi-dimensional (employment) environment it was more / less the same - lots of hype and some anxiety (of computers replacing human labour / function). I'm following the same trend here. During the 1990's while I was contracted to Intel Corporation this became a regular international analytical / strategic marketing 'calibration' on various technology and socio-demographic employment dimensions at the time.
Now, with a mature global IT / Internet environment many new jobs / industries were created - and with the development of contemporary psychological / neuroscientific research more information is available (to a broader audience) for online / offline publication for discussing / predicting the effect of AI on mental health (i.e. anxiety of possible unemployment in certain sectors where it could replace human labour).
The impact of AI application / information (speculation / functionality / advantages / disadvantages) is in my opinion following the same trend as the global IT discussions we had during the 80's / 90's, albeit now with more definite qualitative (i.e. technology / neuro) research, application, functionality, prediction and probability.
AI is now at the same time also touted to exert a substantial future contribution in assisting individuals with various anxiety and depression-related disorders. Many research articles and journals (across various technology, health disciplines and internationally respected academic sources) are currently publishing and proposing various AI-assistance technology-based experiments and mental wellness guidance applications.
I'm of opinion, as with IT in a sense, the human (and employment) will evolve and migrate to an AI-enriched socio-economic society (where applicable). 'AI Anxiety' is perhaps a premature (or overstated) 'diagnosis' of (un)certain future human behaviour or related pathologies.
IT brought with it various behavioral addictions (i.e. online gaming / gambling / Internet) which became classified, diagnosed and treatable contemporary psychological disorders. This in itself created employment for various other job-types (across the global mental health fraternity, for example).
AI could therefore offer similar human behavioral threats (currently or as it evolves into a mature state of application and functionality). I will not be surprised to witness 'AI Anxiety and related disorders' in future additions of i.e. the Diagnostic and Statistical Manual of Mental Disorders (DSM) Editions - as researched, diagnostically classified and published by the American Psychiatric Association (APA).
I, for one, is of firm believe that AI will probably generate many human opportunities and threats - it will depend how individuals, economies, governments and the world at large define and regulate (and / or perhaps in cases mitigate related anxiety / uncertainty) it's value in future human life and employment.
Keep in mind that various associated media publications may also play a role in creating 'AI sensation', especially in a potentially 'disruptive technology" environment orientation.
Vernon Chalmers Disclaimer: Above response is my personal opinion(s).
© Vernon Chalmers : Mental Health and Motivation (on 'AI Anxiety')