二 | 散文《家徽》本是湖北作家胡成中的作品，在网上却长期被说成是余华写的。不只余华，莫言也专门写过文章辟谣，鲁迅、张爱玲、杨绛等已故文学大家就更别提了，网络上充斥着冠了他们名字的伪语录，相信大家都见过鲁迅先生“这不是我说的”表情包。    Asha Mandela from the US state of Florida broke her own Guinness World Record for having the longest dreadlocks, which now measure up to 33.5 meters (110 ft).,Calling her locks "Mr. Cobra" and a "royal crown", Mandela shared that she began growing them over 40 years ago when she moved from the island nation of Trinidad and Tobago to New York.,Interestingly, to maintain and groom her dreadlocks, Asha's husband, a hair stylist, helps her wash them once a week with up to six bottles of shampoo. It takes two days to dry them.,To prevent her dreadlocks from dragging on the ground, Asha carries them in a fabric sling and sometimes wears her locks up on her head, but only for a few hours as her neck begins to strain.,"When I’m ready to go into my sleep chamber with my cobra baby, I would have them tied up in a little sack and we cuddle and talk to each other.",Mandela, who nurtures her locks like a plant, doesn't plan to ever chop them as she feels they have become a part of her.,“I will never see or feel that there’s a time that I want to cut my hair,” Mandela said.。假署名、伪语录之所以层出不穷，背后是低风险、高收益的算计。一些自媒体为博取关注，有的移花接木，直接篡改作品署名，把无名作者置换成知名作家；有的凭空虚构，拼凑、改写甚至编造所谓的“名人金句”。这类操作几乎零成本，只要贴上名人标签，就能更容易收割流量、获得可观的点击与转发，从中牟取商业利益，而受害者取证、辟谣、维权却要耗费大量时间与精力。北京鲁迅博物馆就专门上线检索系统，用来甄别网上五花八门的“鲁迅名言”，但仍然追不上虚假内容的流转速度。

三 | 这些内容一经生成，便能通过一键转发、算法助推产生裂变式传播，传播链条长、覆盖面广，旧的不实链接刚被处理，新的虚假版本又快速生成。当前的AI工具在抓取信息时无法区分真假，部分假内容被大模型抓取、纳入数据训练，经过二次生成后进一步固化，形成“越传越错”的恶性循环。类似“假鲁迅”“假余华”的内容泛滥，伤害的不只是创作者，更波及广大受众与整个网络文化生态。原作者劳动成果被随意挪用，署名权遭到侵害；被冒名者的思想观点、公众形象遭到曲解，无端承受误读与争议；受众难辨真伪，把拼凑杜撰的文字当作名家名句；部分伪作甚至流入出版物、教辅资料和考试试卷，对青少年形成误导。

四 | 凡此种种，不仅对网络文化生态造成侵蚀与破坏，也消解了全社会尊重原创成果、保护知识产权的共识。对此，平台和AI内容服务商责无旁贷，应严格规范作品作者信息标注，建立作品权属核验机制，对涉及知名作家的内容加强筛查审核，防范AI幻觉生成伪名人内容，从源头遏制冒名侵权。

五 | 打假不应是创作者一个人的事，每多一层审核、多一次核实，虚假信息就少一分流传的空间。

六 | 让被假冒的风险、打假的成本从创作者身上卸下来，这才是对原创最大的尊重。来源：人民日报客户端，作者：郑娜，策划：岳小乔、雒健晴，海报：陈泉伊本期编辑：胡洪江、阮可欣