A daily overview of the AI landscape as of July 7, 2025
As the summer heat settles over Europe, the world of AI is showing its dynamic side. Today, Monday, July 7, 2025, important developments are emerging that bring both opportunities and challenges for the future of digital technologies.
Google expands Gemini access to private communications
Google today made a significant change to its AI assistant Gemini. Gemini can now access WhatsApp, Messages, and the Phone app on Android devices—regardless of whether users have enabled Gemini Apps Activity.
This development marks an important step toward "ambient computing," in which AI systems are more closely integrated into everyday communication. Google assures that Gemini cannot read or summarize WhatsApp messages, but is only intended to assist with the use of the apps.
However, the change raises questions about digital privacy. Users can disable the feature in the settings, but they must first be aware that it has been enabled by default.
Microsoft relies on AI infrastructure – with consequences for jobs
Microsoft has announced further job cuts totaling approximately 9,000. The Xbox division is particularly affected, where projects such as the "Perfect Dark" reboot and "Everwild" have been canceled.
These decisions are related to Microsoft's massive investments in AI infrastructure. The company plans to invest around $80 billion in this area in fiscal year 2025.
Phil Spencer, CEO of the Gaming Division, emphasized in an internal statement: "Despite the layoffs, the Xbox platform, hardware, and game roadmap have never been stronger." This statement reflects the challenges associated with the current transformation of the technology industry.
OpenAI specifies GPT-5 plans
OpenAI has confirmed that GPT-5 will definitely be released in summer 2025. The new system will combine the strengths of different model lines for the first time – the multimodality of GPT-4 with the reasoning capabilities of the "o" series.
This announcement comes at a time of increasing competition. DeepSeek from China offers cost-effective alternatives, while Anthropic's Claude is gaining popularity among developers. GPT-5 will have to prove whether it can meet the high expectations.
Change in media usage
A recent study shows significant changes in media consumption: News-related queries to ChatGPT increased by 21% in the US between January 2024 and May 2025, while similar Google searches decreased by 5%.
Organic visits to news sites fell from 2.3 billion to under 1.7 billion. At the same time, ChatGPT referrals to news sites increased 25-fold. Reuters, the New York Post, and Business Insider benefited most from this change.
AI breakthroughs in pharmaceutical research
The company uses AlphaFold 3 to accurately predict protein interactions. With $600 million in funding and partnerships with Novartis and Eli Lilly, AI is emerging as a tool for solving concrete medical challenges.
Robotics forecasts and investment trends
Investor Vinod Khosla predicts a "ChatGPT moment" for robotics in the next two to three years. His vision: "Almost everyone will have a humanoid robot at home by the 2030s, probably starting with specialized tasks like cooking."
The financial markets are already showing clear trends: AI startups received 53% of all global venture capital funding in the first half of 2025—and in the US, even 64%. This capital concentration significantly exceeds historical comparable values.
Alibaba Cloud is accelerating this development with a $60 million investment in its partner network. With 12,000 partners worldwide, including Salesforce, IBM, and Neo4j, a global AI distribution network is emerging. Raymond Ma, VP of Global Partners at Alibaba Cloud, explains: "We're not just building a channel, but a synergistic ecosystem that accelerates digital transformation."
Meanwhile, TSMC is benefiting from high demand for AI chips as Nvidia, AMD, and Apple outsupply – a trend that underscores the hardware-driven AI revolution.
Regulatory debates in the US and Europe
At the regulatory level, clear lines are emerging. The US Senate voted 99-1 against Trump's proposed 10-year moratorium on government AI regulation. In Brussels, the European Commission is sticking to its AI Act timeline despite intensive lobbying from tech companies.
At the same time, Cloudflare is responding to AI companies' data collection mania: AI crawlers are now blocked by default, giving publishers more control. CEO Matthew Prince explains: "The goal is to return power to creators without hindering AI innovation." Over a million websites have already adopted this policy.
Experimental developments at Meta
Meta is currently testing chatbots that spontaneously message users and recall past conversations. These experiments demonstrate the boundaries between helpful AI support and potentially intrusive interaction.
AI adoption in the workplace
The transformation of the world of work through AI is progressing significantly. The Slack Workforce Index shows a remarkable development: AI usage has increased by 233% since November 2024. Employees who use AI daily are 81% happier and 64% more productive.
At the same time, studies show that at least 13.5% of biomedical abstracts show signs of AI support, in some cases as much as 40%. Certain words like "crucial" and "significant" have appeared more frequently since ChatGPT's release—a linguistic fingerprint of the AI revolution.
Paradoxically, Western tech giants like Amazon, Microsoft, and Google are using Chinese AI models like DeepSeek – despite US security concerns. This development demonstrates how pragmatic business interests can overcome geopolitical tensions.
outlook
The coming days will bring further developments: OpenAI is expected to release additional GPT-5 details, while Microsoft must deal with the backlash to its restructuring. Q2 earnings from various AI companies will show whether the high valuations are justified.
The BBC is dedicating an entire program to Artificial General Intelligence tonight, featuring Yoshua Bengio, one of the "Godfathers of AI." While the major players—Google's DeepMind, OpenAI, and Meta—can't yet agree on a definition of AGI, CEOs are confident that AGI will be the next big breakthrough.
At the same time, research from Giessen shows that AI still struggles to recognize what doesn't fit into a sequence—an example of how far we still are from human cognition. While humans intuitively recognize that an elephant doesn't belong with a guitar and a chair, AI often fails to make such categorical distinctions.
Today's developments highlight both the opportunities and challenges of AI transformation. While technological breakthroughs in fields such as medicine promise impressive advances, questions about privacy and jobs raise important societal debates.
The next few months will be crucial for determining the direction in which this development continues – as a tool for solving human problems or as a source of new dependencies.