technology
kē jì
科技
Google
Gǔgē
谷歌
to stop / to prevent
zǔzhǐ
阻止
hacker(s)
hēikè
黑客
to exploit / to make use of
lìyòng
利用
artificial intelligence
rén gōng zhì néng
人工智能
to develop
kāifā
开发
two-factor authentication
shuāngchóng rènzhèng
双重认证
security vulnerability / exploit
lòudòng
漏洞
Google
Gǔgē
谷歌
researchers
yánjiū rényuán
研究人员
to discover / to find
fāxiàn
发现
one / a (measure word + noun)
yī gè
一个
unknown / unidentified
wèizhī
未知
threat actor
wēixié
威胁
to use / to employ
shǐyòng
使用
artificial intelligence
rén gōng zhì néng
人工智能
to create / to manufacture
zhìzào
制造
zero-day exploit
líng rì lòudòng
零日漏洞
to attempt / to try to
shìtú
试图
to bypass / to circumvent
ràoguò
绕过
two-factor authentication
shuāngchóng rènzhèng
双重认证
system
xìtǒng
系统
Google
Gǔgē
谷歌
subsequently / thereafter
suíhòu
随后
with / and
yǔ
与
the said / that (referring to something mentioned)
gāi
该
software
ruǎnjiàn
软件
vendor / supplier
gōngyìngshāng
供应商
to cooperate / to work together
hézuò
合作
quietly / in a low-key manner
qiāoqiāo
悄悄
to patch / to fix (a flaw)
xiūbǔ
修补
(completed action particle)
le
了
this vulnerability / the said exploit
gāi lòudòng
该漏洞
May 12, 2026 at 9:00 AM
Google
Gǔgē
谷歌
Threat Intelligence Group
wēixié qínggāo xiǎozǔ
威胁情报小组
at / on (formal time/location marker)
yú
于
May 11th
wǔ yuè shí yī rì
五月十一日
to disclose / to reveal
pīlù
披露
one / a (measure word + noun)
yī gè
一个
unknown / unidentified
wèizhī
未知
hacker group / threat actor group
hēikè zǔzhī
黑客组织
to exploit / to make use of
lìyòng
利用
artificial intelligence
rén gōng zhì néng
人工智能
to develop
kāifā
开发
(completed action particle)
le
了
the first (of its kind)
shǒu gè
首个
known / documented
yǐ zhī de
已知的
zero-day exploit
líng rì lòudòng
零日漏洞
this vulnerability / the said flaw
gāi lòudòng
该漏洞
to exist in / to be found in
cúnzài yú
存在于
one / a (measure word for products)
yī kuǎn
一款
popular / widely used
liúxíng de
流行的
open-source
kāiyuán
开源
web-based administration tool
wǎngluò guǎnlǐ gōngjù
网络管理工具
in / within
zhōng
中
capable of bypassing
kě ràoguò
可绕过
two-factor authentication
shuāngchóng rènzhèng
双重认证
protection / security measure
bǎohù
保护
researchers
yánjiū rényuán
研究人员
to point out / to note
zhǐchū
指出
this vulnerability / the said flaw
gāi lòudòng
该漏洞
not / by no means (formal negation)
bìngfēi
并非
to originate from / to stem from
yuányú
源于
common / frequently occurring
chángjiàn de
常见的
programming error / coding mistake
biānchéng cuòwù
编程错误
but rather / instead
ér shì
而是
a type of / a kind of
yī zhǒng
一种
high-level / advanced
gāojí
高级
logical flaw / semantic vulnerability
luójí quēxiàn
逻辑缺陷
attacker(s) / threat actor(s)
gōngjī zhě
攻击者
it seems / apparently
sìhū
似乎
to use / to employ
shǐyòng
使用
artificial intelligence
rén gōng zhì néng
人工智能
model / AI model
móxíng
模型
to (indicating purpose)
lái
来
to identify / to detect
shíbié
识别
this flaw / the said defect
gāi quēxiàn
该缺陷
and / furthermore (formal connector)
bìng
并
to turn it (into) / to convert it
jiāng qí
将其
to transform into / to convert into
zhuǎnhuà wéi
转化为
usable / functional
kě yòng de
可用的
exploit / vulnerability exploit code
lòudòng lìyòng
漏洞利用
Google
Gǔgē
谷歌
to state / to indicate
biǎoshì
表示
this malicious script
gāi èyì jiǎoběn
该恶意脚本
to contain / to include
bāohán
包含
a large number of / abundant
dàliàng
大量
annotations / code comments
zhùshì
注释
style / format
fēnggé
风格
with / consistent with
yǔ
与
large language model (LLM)
dà xíng yǔyán móxíng
大型语言模型
the output of / produced by
de shūchū
的输出
highly consistent / strongly matching
gāodù yīzhì
高度一致
Google
Gǔgē
谷歌
subsequently / thereafter
suíhòu
随后
to notify / to inform
tōngzhī
通知
the said software
gāi ruǎnjiàn
该软件
vendor / supplier
gōngyìngshāng
供应商
both parties / the two sides
shuāngfāng
双方
to cooperate / to work together
hézuò
合作
to patch / to fix (a flaw)
xiūbǔ
修补
(completed action particle)
le
了
this vulnerability / the said flaw
gāi lòudòng
该漏洞
causing it to / making it
shǐ qí
使其
unable to / impossible to
wúfǎ
无法
to be exploited on a large scale
bèi dà guīmó lìyòng
被大规模利用
Google
Gǔgē
谷歌
did not / has not (formal negative)
wèi
未
to reveal / to disclose
tòulù
透露
attacker(s)
gōngjī zhě
攻击者
used by / employed by
suǒ yòng
所用
(possessive/modifier particle)
de
的
artificial intelligence
rén gōng zhì néng
人工智能
model / AI model
móxíng
模型
name / designation
míngchēng
名称
only / merely
jǐn
仅
to state / to indicate
biǎoshì
表示
it / its (formal pronoun)
qí
其
most likely / very probably
hěn kěnéng
很可能
is not / by no means
bìngfēi
并非
Google itself / Google-owned
Gǔgē zìjǐ de
谷歌自己的
Gemini (Google AI model)
Gemini
Gemini