OpenAI has announced autonomous reasoning agents capable of creating new software without human intervention. The announcement has become the latest major development in the field of artificial intelligence, demonstrating how rapidly the agentic systems direction is advancing.
At the same time, Anthropic has warned that it is very close to RSI capability — self-improving artificial intelligence. The company, which is soon to go public, according to the Financial Times, expects a low rating from investment agencies.
Against this backdrop, major technology companies have also become embroiled in legal disputes. Meta, Google and TikTok have filed a lawsuit against the California attorney general over a law regulating social networks for teenagers. The New York Times and other media outlets have filed a lawsuit against the Brave browser, accusing it of unlawful content scraping. Meanwhile, the New York Times is reportedly in negotiations to acquire Letterboxd for $330 million.
On the deals and investment front, Palo Alto Networks is negotiating to acquire the AI security startup Clutch for $500 million. Airbnb has announced that it will release a travel AI agent in the first half of 2027. Serve Robotics has launched a new business with the aim of becoming AWS for robots.
Related News
There is also news for users from Meta: American users can now communicate with their Muse AI avatar via video call. However, Meta warns that election-related backlash against data centers could complicate AI expansion.
The energy problems of data centers are also worsening: they have begun using old aircraft engines due to lengthening waiting times for electricity. Google's first data center satellite is already online and writing poems.
It is noteworthy that the attention of most technology companies is gradually shifting from agentic systems to infrastructure — data centers, energy, security. This is no coincidence. As AI models become more powerful, the resources required for their operation become the main constraint, and this is precisely where it will be decided who can move ahead.
The second trend is legal and political pressure. The lawsuit by Meta, Google and TikTok against the California attorney general, as well as the growing number of communities rejecting Flock cameras, show that the issue of public oversight is becoming the subject of concrete decisions rather than abstract discussion.
Against all this, Anthropic's warning about self-improving AI is worth taking not as a sensation but as a matter of timing. How quickly companies will reach that threshold remains unclear, but the preparedness of investors and regulators will be decisive.
