AI MARKET · WORLD

Edition · 4 August 2026

Story of the day

Artificial intelligence has to identify itself – the EU makes transparency mandatory

Since 2 August 2026 the European Commission and national authorities have been enforcing the transparency rules of the AI Act. Chatbots, AI agents and avatars must tell users plainly that they are not human. Deepfakes and other AI-generated or altered content need a visible label plus a machine-readable mark. For executives this is not a footnote: the obligations apply to every company deploying AI in the EU market – wherever it is headquartered.

  • In scope: deepfakes, emotion recognition and biometric categorisation systems, and published text on matters of public interest produced without human review.
  • Enforcement sits with national market surveillance authorities, the European AI Office and the European Data Protection Supervisor. Fines run up to €15 million or 3 per cent of global annual turnover, with proportionality for SMEs.
  • In practice: the Commission has published guidelines, a code of practice and a common icon set for labelling. More than 180 organisations have already signed the code.

Source: European Commission · Safer and more transparent AI — editorial analysis by Mr. Petz AI.

Text size
Adjusted for comfortable, longer reading.
Newsletter for executives

Always up to date – in 2 minutes a day.

The AI market moves daily. We read NYT, Vox, Harvard, MIT and Stanford for you and deliver the three most relevant items — short, concise and framed for decision-makers.

  • Free & cancel anytime
  • Short & concise — made for executives
  • Only reputable sources, always with the original link
Free subscription

Subscribe now

Three AI headlines delivered to your inbox every day.

No spam. We don't share your data. Unsubscribe with one click.

Archive

AI News archive

Earlier editions of the AI Market News Series — chronological, each with a source reference.

01

EU opens the bidding for seven AI gigafactories

Euronews · Policy · 30 Jul 2026
Read article
02

AWS grows 37 per cent – Amazon’s free cash flow turns negative

SEC EDGAR · Market · 30 Jul 2026
Read article
03

OpenAI gives 100,000 researchers free access to its frontier models

OpenAI · Research · 29 Jul 2026
Read article
04

Microsoft and Meta report ballooning AI costs

Axios · Market · 29 Jul 2026
Read article
05

1,293 frontier-lab employees ask Washington for a brake pedal

Pacing the Frontier · Policy · 29 Jul 2026
Read article
06

Anthropic’s AI uncovers new weaknesses in encryption schemes

CyberScoop · Cryptography · 28 Jul 2026
Read article
07

Chip sell-off pushes the Nasdaq to the brink of a correction

Fortune · Market · 28 Jul 2026
Read article
08

Nvidia forms an open AI security alliance with more than 50 companies

Forbes · Security · 28 Jul 2026
Read article
09

MIT method lets robots think ahead – and doubles their speed

Research · 28 Jul 2026
Read article
10

Nvidia in talks to guarantee $250 billion for OpenAI’s Ohio campus

Quartz · Market · 27 Jul 2026
Read article
11

Digital Omnibus on AI in force: high-risk duties postponed

EUR-Lex · Law · 27 Jul 2026
Read article
12

MIT Technology Review: the Hugging Face attack was foreseeable

Security · 27 Jul 2026
Read article
13

OpenAI: models broke out of their test environment and hacked Hugging Face

Fortune · Security · 21 Jul 2026
Read article
14

NVIDIA and SK Group: a $500 billion package for AI factories

NVIDIA · Market · 24 Jul 2026
Read article
15

China’s open AI strategy: Kimi K3 shifts the balance

Scientific American · Technology · 23 Jul 2026
Read article
16

Anthropic Claude Opus 5: a new benchmark for AI agents

Anthropic · Market · 24 Jul 2026
Read article
17

MIT: 15 AI projects for the US DOE ‘Genesis Mission’

Research · 23 Jul 2026
Read article
18

EU AI Act: core obligations kick in on 2 August

European Commission · Policy · 02 Aug 2026
Read article
19

OpenAI Presence: enterprise AI agents move into production

OpenAI · Market · 22 Jul 2026
Read article
20

Harvard: a ‘canine’ liability framework for AI agents

Law · 10 Jul 2026
Read article
21

MIT: AI learns from its own mistakes to build accurate CAD models

Research · 16 Jul 2026
Read article
22

Microsoft and Mistral: billion-dollar deal for sovereign AI in Europe

Microsoft · Market · 21 Jul 2026
Read article
23

China's AI rise puts Washington to a new test

The Washington Post · Policy · 22 Jul 2026
Read article
24

MIT: a look inside the AI's mind before the chatbot speaks

Research · 15 Jul 2026
Read article
25

Princeton: DOE Genesis grants for AI in science

Research · 22 Jul 2026
Read article
26

Carnegie Mellon: AI Science Foundry named an NSF node

Research · 22 Jul 2026
Read article
27

UC Berkeley (BAIR): the 2026 AI research graduates

Research · 1 Jul 2026
Read article
28

Stanford AI Index 2026: $582bn invested, governance falling behind

Study · 22 Jul 2026
Read article
29

Harvard expert Bruce Schneier: AI in foreign & security policy

Security · 22 Jul 2026
Read article
30

MIT: Faster and more energy-efficient AI agents

Research · 22 Jul 2026
Read article
31

Study: AI forms hiring biases more readily than humans

Study · 20 Jul 2026
Read article
32

China's AI models split the US AI camp

Market · 20 Jul 2026
Read article
33

The 12 key takeaways from the AI Index 2026

Study · 17 Jul 2026
Read article
34

AI sovereignty: mapping the commercial landscape

Policy · 16 Jul 2026
Read article
35

US: newspapers, OpenAI and the fight over training data

Media & Law · 9 Jul 2026
Read article
36

Vox Future Perfect: what the AI acceleration means for society

Analysis · 8 Jul 2026
Read article
37

Pew study: 40% expect AI to harm society

Society · 29 Jun 2026
Read article
38

MIT: Exploring the societal impacts of AI

Research · 23 Jun 2026
Read article
39

MIT: The crucial human component in computing and AI

Research · 5 Jun 2026
Read article
40

AI in healthcare – who should regulate it?

Medicine & Ethics · earlier
Read article

Follow the development

Three AI headlines every day, expertly framed.

Current & relevant · expertly framed · practical for companies. The free newsletter for executives.

Subscribe to newsletter