German Firms Burn Billions on AI, Then Spend Saved Time Double-Checking It
Published on 07/20/2026 at 02:44 | Redaktion boerse-global.de
A major chunk of the efficiency gains companies chase with artificial intelligence is being wasted on manually verifying the tool's outputs, according to a new industry survey that casts doubt on the technology's return on investment. The finding comes as businesses across Europe and the US accelerate AI adoption — and as employee representatives warn the technology is being used as a cover for mass redundancies.
The IDC survey of German finance teams found that 28 percent of the time supposedly saved by AI is consumed by validation work. 18 percent of respondents said they spend more than 30 hours per week checking AI-generated data. The root cause: a lack of transparency in many systems. 68 percent of financial decision-makers refuse to use AI tools whose decisions cannot be explained, even when the tools achieve a 99 percent accuracy rate. Some companies said they would accept an 11 percent cost premium for "glass-box" solutions that make reasoning traceable.
Token-Maxxing: Big Budgets, Diminishing Returns
Analysts describe a pattern of "Token-Maxxing" — corporations exhausting their full AI budgets on infrastructure, security and licenses without achieving the productivity gains they expected. The cost of developing large language models ranges from 60 million to 3 billion US dollars, creating intense pressure to show results.
Roger Basler de Roca, an AI specialist, cautioned: "The sheer frequency of usage is not a measure of success." He said real benchmarks should be time saved, output quality and operational cost.
Job Cuts Masquerading as Efficiency
Skeptics argue that many AI projects are window dressing for layoffs already in the pipeline. Max Votek, a market observer at Customertimes, identified this trend. OpenAI chief Sam Altman has also accused corporations of weaponizing the technology to shed staff.
The numbers are stark. For 2026 alone, companies have announced over 100,000 AI-linked job reductions. Oracle plans to cut 21,000 positions, Citigroup 20,000, and Amazon 16,000. Meta and Salesforce have also conducted deep cuts.
In Germany, Volkswagen is at the center of the storm. Management wants to lift its operating margin from 3 percent to 10 percent. The works council fears the loss of up to 100,000 jobs and the closure of plants in Emden, Hannover and Zwickau.
Courts and Regulators Push Back
Legal frameworks are tightening. In February 2025, the EU AI Act came into force, banning emotion-recognition systems in the workplace. Germany's Federal Labour Court sharpened works council rights in 2024, ruling that employee surveillance systems — including AI-driven ones — are subject to co-determination.
In the United States, the California Nurses Association has criticized Kaiser Permanente for using AI tools that monitor nurses' empathy levels and conversation duration.
A Chinese court in Hangzhou ruled in April 2026 that reliance on AI alone cannot justify a dismissal if the employer did not first examine alternative roles for the worker.
Resistance is building. On 21 July 2026, experts from the German Economic Institute and the Steel Employers' Association will debate the future of social partnership. Simultaneously, IG Metall and the German Trade Union Federation (DGB) are staging protests in Lüdenscheid against job cuts and social austerity.
Disclaimer regarding our articles: No investment advice, no buy or sell recommendation. Information on prices, companies, and markets is provided without guarantee; changes are possible at any time. Stock market transactions can lead to substantial losses. Our articles are created and reviewed in whole or in part automatically with the support of AI.
