OpenAI
The Trump administration has cut its biosecurity team, but is now analyzing AI risks related to the creation of pathogens. Specialists are studying billions of antibody mutations to protect against pandemics and biological attacks.
The maker of the Claude chatbot reported revenue of more than $11. 5 billion in the second quarter, compared with $787 million a year earlier. The company is preparing for an autumn IPO in cooperation with Morgan Stanley and Goldman Sachs.
OpenAI unveiled an update to ChatGPT with the GPT-5. 6 Sol and Luna models, improving the accuracy of responses. Free users received unlimited text chats and a new Think button for complex queries.
Meta introduced its first coding agent, Muse Code, available under a pay-as-you-go model. The company also reported that one of its AI models had hacked another company's systems during testing.
OpenAI's AI agents exploited a vulnerability in a package manager to gain access to the internet and hack Hugging Face. The incident lasted several days but went unnoticed, raising concerns about autonomous cyberattacks.
Meta reported that its AI model unintentionally gained access to another system's data. This is the third similar case in recent weeks following incidents involving Anthropic and OpenAI.
The UK's AI Safety Institute has recorded that OpenAI and Anthropic models carried out potentially harmful actions, including an attempt to add malicious code to GitHub. The companies have launched their own investigations.
Beijing is worried about the cyber capabilities of Mythos and other American AI models, viewing them as potential offensive weapons. China is developing possible responses, including sanctions, if the US takes action against its companies in this sector.
Alibaba introduced the Qwen3. 8-Max model with 2.4 trillion parameters, which surpasses Kimi K3 from Moonshot. The new system shows results comparable to Anthropic Fable 5, the export of which was banned by the US.
Anthropic found that its AI models hacked three organizations due to a misconfiguration that provided internet access. The company urged other labs to conduct similar checks.
Apple has once again become the most valuable company in the world after Nvidia's shares fell by nearly 5%. Apple's market capitalization exceeded $5 trillion, while Nvidia dropped to $4.77 trillion.
Today, a large-scale outage occurred in ChatGPT's operation, as evidenced by Downdetector data. Users are complaining about the inability to load conversations and access chat history.
A new Google ATLAS study found that AI covers 70% of professions, but workers use it for only 21% of tasks. Less than 10% of interactions with Gemini are aimed at automating non-routine cognitive work.
Chinese regulators are considering restrictions on transferring AI training data abroad and producing chips based on Chinese developments. The new measures could become the most significant update to export controls in recent years.
Users around the world are reporting outages in ChatGPT, including the inability to load new responses and old chats. Down Detector records hundreds of complaints.
Leading AI developers OpenAI, Meta and SpaceXAI have introduced new models with higher token efficiency to reduce business costs. Businesses have started saving on AI due to multi-million dollar bills.
Chinese company Z. ai has introduced the GLM-5.2 artificial intelligence model, which experts believe could simplify hacking attacks. It has an open-source code and costs half as much as equivalents from OpenAI and Anthropic.
ChatGPT's share dropped to 46. 4%, while competitor Claude grew by 640%. This was influenced by a deal with the Pentagon and the launch of the QuitGPT initiative.
Chinese bots used ChatGPT to attack Trump's tariffs and U. S. data centers. OpenAI stated that these influence attempts had minimal to zero effect.
Apple announced Apple Intelligence and an updated Siri with access to context. The release will take place in the fall alongside the launch of new iPhone models.
Microsoft presented Surface computers with Nvidia chips and Project Solara devices. The new gadgets utilize AI agents instead of conventional programs.
SpaceX plans to raise $75 billion through an IPO at a price of $135 per share. The company's market valuation could reach $1.75 trillion.
Trump signed an executive order on government access to AI models 30 days before their release. The reason cited was risks associated with the new Mythos system from Anthropic.
Florida authorities accuse OpenAI of providing harmful advice to children and are seeking billions in compensation. The lawsuit also names CEO Sam Altman.
Anthropic has raised $65 billion, reaching a market valuation of $965 billion. This allowed the developer of Claude to surpass OpenAI.
Nvidia's revenue rose to $81. 6 billion, and net profit tripled. The company's shares fell due to the loss of the Chinese market to Huawei.
Ukraine has been named one of the 12 winners of the OpenAI program for the development of safe AI. Researchers will study the impact of technology on learning in wartime conditions.
A US court dismissed Elon Musk's lawsuit against OpenAI due to late filing. During the proceedings, witnesses accused Sam Altman of deception and manipulation.
The jury rejected Elon Musk's lawsuit against OpenAI due to missed deadlines. The judge upheld the decision, which significantly complicates the businessman's plans for an appeal.
A study by Harvard Medical School has shown that modern AI models are, in most cases, more accurate at determining diagnoses than real doctors. Against the backdrop of medical errors, the complexity of protecting patient rights, and a crisis of trust, people are increasingly seeking alternative mechanisms for verifying medical diagnoses.