A late-night live stream organized by AikerWorld and HackingClub on August 4 attracted nearly 8000 viewers and featured nine prominent Chinese security figures discussing how artificial intelligence is reshaping the white-hat ecosystem.
The event was triggered by two high-profile developments: the sudden closure of a long-running financial-sector SRC and HackerOne’s decision to replace anonymous submissions with mandatory real-name registration.
Host Hu Xiaona, 360 Group senior vice president and general manager of VulnCloud, opened the discussion by framing the SRC shutdown not as an isolated incident but as part of a broader structural shift. She noted that AI can now mass-produce vulnerabilities, driving down prices and making it harder for traditional bounty hunters to monetize routine findings.
Hu stated she has already begun guiding thousands of white hats toward AI Forward Deployed Engineer (FDE) positions, describing the role as the “last mile” that helps enterprises actually deploy and secure AI agents in production environments.
Xiong Yong, co-founder of the Hunan Cybersecurity Base, countered that AI pressure is ultimately positive because it forces organizations to treat security as a necessity rather than a compliance checkbox. He urged white hats to move beyond pure offensive work and focus on defensive engineering.
Wei Yongqiang of Kingsoft Cloud warned that devaluing white-hat submissions could push researchers toward gray or black markets, while Luo Xiong of 360 VulnCloud highlighted the difficulty of commercializing AI-security products without strong branding or sales channels.
Additional speakers who joined mid-stream included Yang Wei of ZhongAn Tianxia, external-enterprise practitioner “No Sleep Wind,” state-owned enterprise operator “Mu Zi Li,” Knownsec veteran “Black Brother,” and Hillstone Networks senior vice president Jia Yu. They collectively stressed that AI eliminates low-value, repetitive submissions, raises the bar for high-impact findings, and requires rigorous human oversight to prevent automated tools from causing production incidents such as accidental database deletion.
The panel reached consensus that the old SRC model is evolving from open crowdsourcing toward curated, real-name, invitation-only programs, and that sustainable careers now depend on combining security expertise with AI deployment and business-context skills rather than relying solely on vulnerability bounties.
