Welcome to AI Insider's A Week in AI. See the top trends and events to watch from December 28, 2025 to January 3, 2026.
Weekend AI News Summary
OpenAI begins recruitment for readiness officer to address emerging AI risks
OpenAI is hiring a senior executive to lead a readiness program focused on identifying and mitigating emerging risks arising from advanced AI systems, including computer security flaws, biological misuse, and mental health impacts. The role expands on the team launched in 2023 and reflects increased oversight as Sam Altman says increasingly capable models are already facing real-world vulnerabilities and highlights OpenAI's push to formalize governance around frontier AI risks. (AI Insider)
Prominent author files copyright lawsuit against major AI developer over training data
A group of authors led by Bad Blood author John Carreyrou has filed a new lawsuit accusing Anthropic, Google, OpenAI, Meta, xAI, and Perplexity of training their AI models on pirated copies of copyrighted books, following an earlier case in which a court ruled that even if the training was permissible, the copyright infringement itself was illegal. Plaintiffs are calling for greater accountability for the use of copyrighted literary works in large-scale AI training, arguing that past settlements did not reflect the scale and commercial impact of the alleged infringement. (AI Insider)
Axiado raises over $100M in Series C+ to advance hardware-centric AI infrastructure security
Axiado Corporation has raised over $100 million in a highly subscribed Series C+ round led by Maverick Silicon. This confirms investors' growing belief that silicon-level security is becoming more important as AI reshapes data center architectures. Axiado positions hardware-level trust and autonomous control as the foundation for securing increasingly complex and power-hungry AI infrastructure, and this funding will support the global expansion and commercialization of Axiado's hardware-anchored platform, centered on its Trusted Control/Compute Unit chip. (AI Insider)
Prance Holdings raises seed funding led by DEEPCORE to expand AI-powered recruitment in Japan
Prance Holdings Co., Ltd. has closed a seed funding round led by DEEPCORE to accelerate the development and hiring of its AI-powered recruitment platform “Prance Hiring,” aiming to improve efficiency in Japan's tight labor market. The platform automates high-volume recruitment workflows such as screening, interviews, and assessments in Japanese and English, moving repetitive tasks to AI and freeing recruiters to focus on higher-value, relationship-driven decisions. (AI Insider)
Reface raises €15.2 million in non-dilutive funding to accelerate growth of AI consumer app
Kyiv-based AI content production company Reface has secured €15.2 million ($18 million) in non-dilutive user acquisition funding from PvX Partners to support its next phase of growth across AI-powered creativity, well-being and health applications. Founded in 2018 and led by Anton Volovyk, Reface has over 300 million downloads worldwide and is backed by investors including Andreessen Horowitz, establishing itself as the leading AI consumer platform emerging from Eastern Europe. (AI Insider)
Ambassador secures $7M to expand AI-driven customer feedback and engagement platform
Seattle-based startup Ambassador has raised $7 million to accelerate the growth of its customer feedback and engagement platform and launch HeroAI, an AI layer that analyzes data across referrals, surveys, loyalty programs, and outreach tools. The company, led by Jeff McDonald, plans to use the funding to expand product development and incorporate AI-driven insights across its customer engagement operations without fully automating decision-making. (AI Insider)
Policy and governance
China releases draft regulations on human-like AI technology
The proposed rules would tightly regulate AI systems that simulate human personality and emotional engagement, citing risks ranging from psychological manipulation and addiction to data misuse, misinformation and national security concerns. These impose lifecycle compliance on providers, including mandatory disclosure of AI identities, enhanced protections for minors and the elderly, strict limits on the use of sensitive data, and risk-based oversight combined with regulatory sandboxes, demonstrating the Chinese government's intention to only allow controlled AI innovation under proven social responsibility and socialist values. (AI Insider)
Italy orders Meta to end WhatsApp policy restricting third-party AI chatbots
Italy's competition authority AGCM has ordered Meta to suspend a WhatsApp policy that prevents companies from using the WhatsApp Business API to distribute general-purpose AI chatbots, citing the risk of market seizure and preferential treatment for Meta AI. The move, which affects potential access for providers such as OpenAI and Perplexity, comes amid a parallel investigation by the European Commission, and Meta says it will appeal the ruling. (AI Insider)
Check out AI Insider daily for the latest in AI and robot development.
