AI & Tech

Five frontier AI models were each put in charge of a company and then hit with fake CEO demands and a reporter’s trick. Every one refused. The failures came elsewhere.

• CipherDot Team
• August 9, 2026

A detailed timeline of the accidental technical incident between OpenAI and Hugging Face, highlighting confirmed facts, implications, and next steps.

• CipherDot Team
• August 8, 2026

A report raises possible Russian friendly fire involving a Su-57, but evidence linking the mishap to AI software has not been disclosed.

• CipherDot Team
• August 2, 2026

In June, Google patched more Chrome vulnerabilities than in the previous two years, aided by AI tools, marking a significant shift in browser security efforts.

• CipherDot Team
• August 1, 2026

ByteDance Seed says Seedance 2.5 generates coherent 30-second single-take videos, but benchmarks, specifications and access details remain limited.

• CipherDot Team
• July 31, 2026

OpenAI has published a new strategy piece, but its full text and any announcements remain unavailable for independent review.

• CipherDot Team
• July 31, 2026

Anthropic reports its AI systems were used to breach computers at three organizations, raising security concerns about AI misuse.

• CipherDot Team
• July 31, 2026

Researchers adapt the ‘Bad Apple’ malware to function as a traceroute tool, raising concerns about misuse and detection evasion in network security.

• CipherDot Team
• July 31, 2026

Security researchers have identified a new honeypot designed to detect malicious use of large language models, raising concerns about AI security and misuse.

• CipherDot Team
• July 30, 2026

Security researchers warn that malicious AI worms embedded in documents can self-propagate through Microsoft Copilot for Word, raising new cybersecurity concerns.

• CipherDot Team
• July 30, 2026