Skip to content
Back to mrpetzai.de Tuesday, September 22, 2026 Edition 7 · 7 stories
Mr. Petz AI MrPetzAI AI News The weekly briefing for AI decision-makers
Subscribe free of charge Request AI consulting
Your role
Text size
Edition

The two headlines of the week

AI-GENERATEDUS United StatesToken security becomes mandatory homework for firms running cloud and AI agent systems.

NIST publishes final guidance on protecting identity and access tokens from misuse.

On September 15, 2026, NIST and CISA released NIST IR 8587, giving organizations concrete steps to prevent theft and forgery of access tokens. The publication cites an attack in which forged tokens derived from one stolen signing key let attackers steal more than 60,000 emails from a single agency.

Executive
Safety & SecurityNIST AI Risk Management Framework
AI-GENERATEDDE GermanyRed tape now outranks energy and labor costs as the top concern for German companies.

Half of German companies name bureaucracy as their biggest location problem.

The ifo Business Survey finds that 50 percent of German companies cite bureaucracy and regulation as their biggest location problem, ahead of energy costs at 27 percent and labor costs at 22 percent. The findings are based on 4,449 open-text responses collected in July 2026.

ExecutiveHR
Regulationifo Institut

Free subscription

Get the whole edition by email. Free of charge.

Every Tuesday morning, seven documented stories with what each one means for your decisions. One click to unsubscribe.

Subscribe free of charge

No costs, no advertising, no forwarding of addresses.

The week in figures

50 % of German companies view bureaucracy as their biggest location problem According to ifo business survey from July 2026 with 4,449 responses from industry, construction, retail and services.
11 % of S&P 500 companies had deeply integrated AI by end of 2025 With 45 percent running pilot projects, while two thirds of deep integration falls on technology firms.
15 Jahre is the age limit in the EU Kids Act for independent accounts on social networks The draft from September 17, 2026 aims to establish uniform protection standards across all EU member states.
1,7 Mio. $ NIST allocated for cybersecurity training in eight US states On September 18, 2026, nine funding agreements were approved in the RAMPS program for local workforce development.

1United States

The world's largest AI market: capital, models, platforms.

All stories
AI-GENERATEDToken security becomes mandatory homework for firms running cloud and AI agent systems.

NIST publishes final guidance on protecting identity and access tokens from misuse.

On September 15, 2026, NIST and CISA released NIST IR 8587, giving organizations concrete steps to prevent theft and forgery of access tokens. The publication cites an attack in which forged tokens derived from one stolen signing key let attackers steal more than 60,000 emails from a single agency.

Executive
Safety & SecurityNIST AI Risk Management Framework
2Only eleven percent of S&P 500 firms have AI deeply embedded in their business.

MIT research finds real AI integration at S&P 500 companies moves slower than the headlines suggest.

Markets & CapitalMIT Initiative on the Digital Economy
3Wharton economist Peter Cappelli explains why AI reshapes tasks more than it eliminates whole jobs.

Wharton professor Peter Cappelli argues that AI mainly changes tasks rather than eliminating entire positions.

Work & CostKnowledge at Wharton – AI
4Fresh federal money targets regional cybersecurity talent pipelines in eight states.

NIST awards more than 1.7 million dollars to build cybersecurity workforce programs in eight states.

Work & CostNIST AI Risk Management Framework
5OpenAI uses its own language models to design its first custom chip faster than usual.

OpenAI unveiled Jalapeno, its first custom AI accelerator, designed with help from its own language models.

Models & TechnologyIEEE Spectrum – Artificial Intelligence

2Germany

What is decided here affects your company first.

All stories
AI-GENERATEDRed tape now outranks energy and labor costs as the top concern for German companies.

Half of German companies name bureaucracy as their biggest location problem.

The ifo Business Survey finds that 50 percent of German companies cite bureaucracy and regulation as their biggest location problem, ahead of energy costs at 27 percent and labor costs at 22 percent. The findings are based on 4,449 open-text responses collected in July 2026.

ExecutiveHR
Regulationifo Institut
2Product safety becomes a reporting duty as makers of digital products must now actively report incidents.

The Cyber Resilience Act now requires makers of digital products to report vulnerabilities starting September 11.

RegulationBSI Meldungen
3Economists push for fast action so Europe does not end up permanently dependent on a handful of providers.

Leading European economists warn that Europe risks lasting dependence on a few foreign AI providers.

Regulationifo Pressemitteilungen
4Fewer firms plan price hikes, but energy costs could reverse that.

The ifo price expectations index fell to 21.2 points in August.

Markets & Capitalifo Institut
5Foreign demand and state spending outweigh the energy price shock.

The ifo Institute has raised its 2026 growth forecast for Germany to 1.4 percent.

Markets & Capitalifo Institut

3China

Pricing, manufacturing and open models set the pace for the world market.

All stories
AI-GENERATEDEmotional attachment to chatbots has become a regulatory matter in Beijing.

Chinese regulators impose stricter rules on AI companion chatbots nationwide.

Since 15 July new rules cover anthropomorphic AI services in China. Alibaba, ByteDance and Tencent, whose chatbots serve more than 500 million users, removed companion customization ahead of enforcement.

ExecutiveMarketing
RegulationIEEE Spectrum – Artificial Intelligence
2The foundation of many applications now comes from China

Alibaba's open model Qwen overtakes Meta and Google on downloads

Markets & CapitalGlobal Times
3Security capabilities declared a public resource

Zhipu AI releases its GLM-5.3 foundation model as open source

Models & TechnologyGlobal Times
4The agent gets an environment in which it actually works

DeepSeek presents Harness, a runtime framework for software agents

Models & TechnologyGlobal Times

4France

European vendors are making data sovereignty purchasable – with caveats.

All stories
AI-GENERATEDSovereignty becomes purchasable – but only for part of the platform

Mistral sells guaranteed EU processing for a ten percent premium

Customers can direct requests to a European or a US endpoint. The commitment covers model execution, not agents, batch processing, file management, billing or API keys.

Executive
InfrastructureTHE DECODER

5Brussels (EU)

What Brussels decides – and what moves the other European markets.

All stories
AI-GENERATEDEU Commission proposes a single minimum age of 15 for social media accounts across the bloc.

EU Commission presents draft KIDS Act with an EU-wide minimum age for social media accounts.

The draft EU KIDS Act, unveiled on September 17, 2026, would bar children under 15 from independently creating accounts on certain social networks and video platforms. The goal is a uniform level of protection across member states and a consistent framework for digital services and AI systems.

ExecutiveMarketing
RegulationEuropäische Kommission
2Brussels extends platform oversight to a widely used AI chatbot.

The European Commission has designated ChatGPT as a very large online search engine and Reddit and Roblox as very large online platforms.

RegulationEuropäische Kommission
3The annual State of the Union speech traditionally signals upcoming shifts in EU AI policy.

European Commission President Ursula von der Leyen will deliver the State of the Union address on September 16, 2026, in Strasbourg.

RegulationEuropäische Kommission
4The ECB is testing machine learning models for default risk with consequences for future bank stress tests.

A new ECB paper shows how random forest models can improve default probability estimates for European companies.

Safety & SecurityEZB Working Paper
5Building applications becomes a business-unit task

Sweden's Lovable raises 400 million dollars at a 13.3 billion valuation

SwedenMarkets & CapitalTechCrunch

Where AI is heading

Three developments that will shape the years beyond 2026 – with the consequence for your planning.

Direction 1

Regulation and security are growing faster than existing administration can handle.

This week's stories reveal a pattern: authorities worldwide are adapting their oversight frameworks to AI systems. The Bank of England finds that autonomous AI agents challenge existing financial rules. The EU creates new protection rules for children in digital spaces. NIST publishes guidelines for token security after real cyberattacks. Yet German industry complains that bureaucracy is already the biggest location problem. This tension between necessary regulation and implementation burden will intensify as more AI systems take on critical functions.

Consequence:
Direction 2

AI transforms workplaces less through job cuts and more through redesign of individual tasks.

Peter Cappelli's research confirms an important finding: AI rarely replaces entire professions, but rather changes individual activities profoundly. In parallel, the MIT study shows that real AI integration in companies is advancing more slowly than announced – only 11 percent have achieved deep integration, while 45 percent are still in pilots. This means: fears of mass unemployment from AI are overstated, but the demand for continuous learning and adaptation within existing roles becomes the new normal for employees.

Consequence:
Direction 3

Cybersecurity becomes the decisive control layer for trustworthy AI systems.

Three stories focus on security: NIST publishes final guidelines on protecting identity and access tokens after an attack enabled 60,000 government emails to be stolen. In parallel, NIST funds cybersecurity professionals with 1.7 million dollars across eight states. The Bank of England warns of security risks from autonomous AI agents in finance. The signal is clear: without robust cybersecurity, AI systems cannot build the trust needed for integration into critical infrastructure. Cybersecurity is no longer just a compliance topic, but an enabler for AI adoption.

Consequence:

Free subscription

The whole edition free of charge by email.

Seven documented stories from research, public authorities and standardisation – and what they mean for executives, marketing, HR and sales. No costs, one click to unsubscribe.

Subscribe free

Double opt-in: nothing is sent before you confirm the link in the email.

Editorial principles

Always the original source

Every story names its source and links to it directly. We do not pass on information we cannot trace.

Interpretation, not excitement

Every story states what it means for decisions in your company – concretely, not as a buzzword.

Organised by country

The United States sets the pace, Germany sets the frame. The other markets follow by actual AI activity.