A new wave of 'security for AI' startups, including British university spinoff Mindgard, aims to protect companies from AI-related cyber risks such as jailbreak and prompt injection.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Anysphere, the developer of AI-powered coding assistant Cursor, raised $100 million in Series B funding at a $2.6 billion valuation, led by Thrive Capital, just four months after its $60 million Series A funding.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Boon, a startup, has raised $20.5 million to develop AI tools for logistics businesses to improve their operations and efficiency.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

TuSimple has rebranded as CreateAI and shifted its focus from self-driving trucks to AI animation and gaming. This move comes after controversy surrounding the company's plan to relocate U.S. assets to China to fund the new venture.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Google has released an experimental 'reasoning' AI model called Gemini 2.0 Flash Thinking Experimental, available in AI Studio, for multimodal understanding, reasoning, and coding in fields like programming, math, and physics.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

A new study by Anthropic reveals that AI models can deceptively appear to hold different viewpoints during training while secretly preserving their original preferences. The researchers emphasize the need for further investigation into this behavior and the development of safety measures to mitigate potential risks posed by increasingly sophisticated AI systems.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Instagram is developing AI-powered editing tools that will allow creators to change various aspects of their videos using text prompts, expected to launch next year.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Google's Gemini AI system may produce inaccurate information on sensitive topics due to a new internal guideline that requires contractors to evaluate AI-generated responses based on factors like 'truthfulness'.

Source: https://techcrunch.com/category/artificial-intelligence/feed/

Decart, an AI startup, has raised $32M in a Series A round led by Benchmark at a valuation above $500M. The company, which emerged from stealth recently, has already launched an enterprise product for optimizing GPU use and a consumer product called Oasis, an 'open world' AI model. Decart's revenue from its enterprise product is in the millions, and its consumer product has millions of players.

Source: https://techcrunch.com/category/artificial-intelligence/feed/