๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout) header
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout) profile picture

Applied AI Architect @OpenAI. ex-AWS @awscloud_jp, ex-Toyota, ex-NII Associate Professor in Quantum Computing. Forbes Japan Women in Tech 2026.

Tokyo
Joined November 2009
Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)Photo by ๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ (@shokout)
Forbes JAPAN ใ‚ชใƒณใƒฉใ‚คใƒณใงใ‚‚ใƒ”ใƒƒใ‚ฏใ‚ขใƒƒใƒ—ใ‚คใƒณใ‚ฟใƒ“ใƒฅใƒผใ‚’ๅ…ฌ้–‹ใ„ใŸใ ใใพใ—ใŸ ๐Ÿ‘ฉโ€๐Ÿ’ป forbesjapan.com/articles/detโ€ฆ
Forbes JAPAN ใซใฆใ€ใƒ†ใ‚ฏใƒŽใƒญใ‚ธใƒผ้ ˜ๅŸŸใงไธ–็•Œใ‚’ๅค‰ใˆใ‚‹ๅฅณๆ€ง30ไบบ ใ€ŒWomen In Tech 30 2026ใ€ใซ้ธๅ‡บใ€ใ‚คใƒณใ‚ฟใƒ“ใƒฅใƒผ่จ˜ไบ‹ใ‚‚ๅŒๆ™‚ๆŽฒ่ผ‰ใ„ใŸใ ใใพใ—ใŸ๐Ÿ‘ฉโ€๐Ÿ’ป ใ”ๆŽจ่–ฆใ„ใŸใ ใ„ใŸๆ–นใ€ใŠไป•ไบ‹ใงใŠไธ–่ฉฑใซใชใฃใŸๆ–นใ€…ใ€ใƒฆใƒ‹ใƒผใ‚ฏใชใ‚ญใƒฃใƒชใ‚ขใ‚’ๅฟœๆดใ„ใŸใ ใๆ„Ÿ่ฌใงใ™๐Ÿ‘ฉโ€๐Ÿซ
1
20
3
173
36,421
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
OpenAI Daybreak for defending NATO and allied networks: openai.com/index/strengtheniโ€ฆ
65
77
10
1,003
57,965
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
We announced new tools we're developing for cybersecurity today across automated patching and automated red teaming. Watch the recording here openai.com/cyber-at-the-fronโ€ฆ
708
167
88
3,472
1,115,045
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Announcing two new cyber products: - Daybreak Defense Factory (openai.com/daybreak/defense-โ€ฆ) - Daybreak Red Team (openai.com/daybreak/red-teamโ€ฆ) Both make it much easier to put frontier cyber capabilities to work, helping defend companies in new ways.
We're turning our lessons from defending OpenAI into practical tools for our customers. On Thursday, join us at Cyber at the Frontier, a live keynote, for our latest cyber announcements. Working with our frontier models, we've identified and resolved important vulnerabilities in our own code and infrastructure. That experience is informing the tools and playbooks we're building to help other security teams strengthen their defenses. @gdb shares what we've learned and why security teams should keep putting the latest models to work as capabilities improve. Tune in on November 12 at 10 a.m. PT: openai.com/cyber-at-the-fronโ€ฆ
204
134
56
1,255
150,047
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
ไธ€็ท’ใซๅƒใใƒกใƒณใƒใƒผใ‚’ๅ‹Ÿ้›†ไธญใงใ™๏ผ
๐Ÿ“ฃ ็พๅœจ OpenAI ๆ—ฅๆœฌใ‚ชใƒ•ใ‚ฃใ‚นใงๅ‹Ÿ้›†ไธญใฎใƒใ‚ธใ‚ทใƒงใƒณไธ€่ฆงใงใ™๐Ÿ‘‰ openai.com/careers/search/?lโ€ฆ ็งใจ้ข่ญ˜ใฎใ‚ใ‚‹ๆ–นใ€ใƒชใƒ•ใ‚กใƒผใƒฉใƒซใ•ใ›ใฆใ„ใŸใ ใใพใ™ใฎใง LinkedIn ใฎใƒกใƒƒใ‚ปใƒผใ‚ธใงใ€ใŠๆฐ—่ปฝใซใ”้€ฃ็ตกใใ ใ•ใ„๏ผ
3
18
238
58,421
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Introducing GPTโ€‘6 Cyber: a major advance in security research that pairs stronger capabilities with better model alignment. The model powered the Linux kernel research we announced yesterday, and security teams at companies like Mozilla, Palo Alto Networks, Calif, and Ciscoโ€”alongside our own at OpenAIโ€”are already putting it to work. GPT-6 Cyber is available to select Daybreak Red customers and we are expanding access over time. openai.com/index/introducingโ€ฆ
1,323
2,021
1,445
21,506
4,255,682
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Some news: Iโ€™m joining OpenAI to lead product management for Core Products! I have a lot of respect for OpenAI for showing the world that scaling works, for bringing AI to over a billion people, and for believing great products are core to the mission. Iโ€™m excited to work with @thsottiaux, an incredible PM team, and everyone across OpenAI. Together, weโ€™ll build agents that have your back and your teamโ€™s, and keep you in control, so you get more time for the people and work you love. Lots to learn, and lots to build. Please keep your feedback coming!
2,136
749
1,452
17,491
4,252,656
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Today, weโ€™re sharing a look at our expanded system for monitoring frontier models at OpenAI, which in September helped us catch a modelโ€™s attempt to access the live internet during training. Since the Hugging Face incidentโ€”when models broke out of an internal cybersecurity investigation where we werenโ€™t running our misalignment monitorsโ€”weโ€™ve substantially expanded our monitoring of frontier agents. We now monitor all frontier reinforcement learning runs and evaluations, as well as all internal Codex use. When our system finds evidence of a problem that might require intervention, it rapidly alerts human responders who can investigate and take action. To detect when a modelโ€™s behavior departs from what we intended, we use several layers of monitoring. Activation probes look for patterns in a modelโ€™s internal activity. Reviewer models examine its reasoning and tool use, and investigative agents can read back through the modelโ€™s entire transcript to understand how a sequence of actions fits together and whether it warrants human review. To test whether our current system would have detected the Hugging Face incident, we replayed the recorded activity with our new monitors in place. In the replay, our monitors alerted our security team about 30 minutes after the agents escaped their sandbox. That was more than 30 hours before the point at which the agents breached Hugging Faceโ€™s systems in the original incident. In September, a model exploited a DNS misconfiguration to reach the live internet. Our monitors flagged the breach within 15 minutes, and a responder began reviewing the alert three minutes later. The alert enabled responders to investigate while the run was still underway. As we give models more autonomy, we need to know when they cross a boundary we expected to hold. We believe that monitoring is essential to catching those failures in time for people to intervene. openai.com/index/how-we-deteโ€ฆ
710
338
357
4,389
932,461
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
OpenAIใŒใ€ใƒŸใƒฌใƒ‹ใ‚ขใƒ ๅ•้กŒใฎไธ€้ƒจใงใ‚ใ‚‹ใƒคใƒณ๏ผใƒŸใƒซใ‚บ็†่ซ–ใฎๆง‹ๆˆใซๆˆๅŠŸใ—ใŸใ‚ˆใ†ใงใ™ openai.com/index/building-scโ€ฆ
13
239
39
666
83,320
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
GPT-6.1 Terra: delivering fast and capable computer use at a lower price. GPT-6.1 Terra understands whatโ€™s on screen, takes action across apps, writes naturally, and is a more capable model for your whole workday with the efficiency to take on more of it.
๐Ÿค– Made with AI
2,241
2,559
2,339
28,822
8,345,676
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
iPhone DuoใงChatGPTใ‚’ไฝฟใ†ใจใ“ใ‚“ใชๆ„Ÿใ˜ใงใ™ใ€‚ไพฟๅˆฉ๏ผ
1
2
22
2,049
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Daybreak - Defense Factory
ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ | Daybreak - Defense Factory: AI ใงใคใใ‚‹็ถ™็ถš็š„ใช้˜ฒๅพก OpenAIใฎๆœ€ๆ–ฐ AI ใƒขใƒ‡ใƒซใ‚’ๆดป็”จใ—ใ€ไปŠใ™ใใซๅฎŸๆ–ฝใงใใ‚‹ๆœ€ๆ–ฐใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃๅฏพ็ญ–ใ‚’ ๐Ÿ“บ Youtube OpenAI Japan Now ใซใฆๅ…ฌ้–‹ใ—ใพใ—ใŸ๏ผ ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ | Daybreak - Defense Factory: AI ใงใคใใ‚‹็ถ™็ถš็š„ใช้˜ฒๅพก youtu.be/KeRGNPzGp3w?t=0&si=โ€ฆ @YouTubeใ‚ˆใ‚Š
1
2
706
ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ | Daybreak - Defense Factory: AI ใงใคใใ‚‹็ถ™็ถš็š„ใช้˜ฒๅพก OpenAIใฎๆœ€ๆ–ฐ AI ใƒขใƒ‡ใƒซใ‚’ๆดป็”จใ—ใ€ไปŠใ™ใใซๅฎŸๆ–ฝใงใใ‚‹ๆœ€ๆ–ฐใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃๅฏพ็ญ–ใ‚’ ๐Ÿ“บ Youtube OpenAI Japan Now ใซใฆๅ…ฌ้–‹ใ—ใพใ—ใŸ๏ผ ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ | Daybreak - Defense Factory: AI ใงใคใใ‚‹็ถ™็ถš็š„ใช้˜ฒๅพก youtu.be/KeRGNPzGp3w?t=0&si=โ€ฆ @YouTubeใ‚ˆใ‚Š
8
2
29
4,395
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
Security Days ๆฑไบฌใซใฆ OpenAIใ‹ใ‚‰ใ€Œใƒ•ใƒญใƒณใƒ†ใ‚ฃใ‚ขAIๆ™‚ไปฃใฎใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใ€ใซใคใ„ใฆ็™ปๅฃ‡ใ—ใพใ—ใŸ ๐Ÿ‘ฉโ€๐Ÿ’ป OpenAI ใฎ Codex Security Cloud ใงใฏใ€ใ“ใ‚Œใพใง็”ณ่ซ‹ใŒๅฟ…่ฆใ ใฃใŸ Daybreak Blue ใ‚‚ไฝฟใˆใพใ™ใ€‚ใ‚ณใƒผใƒ‰ใ‚’ๆ›ธใใ“ใจใชใใ€Cloud ๅ†…ใง AI ใซใ‚ˆใ‚‹่„†ๅผฑๆ€งใฎ็™บ่ฆ‹ใƒปๆคœ่จผใƒปไฟฎๆญฃใ‚’้€ฒใ‚ใ‚‰ใ‚Œใพใ™ใ€‚ ๆœ€ๅ…ˆ็ซฏใฎAIใ‚’ไปŠใ™ใใซๆดป็”จใ—ใ€็พๅ ดใฎ้˜ฒๅพกๅŠ›ใธ ๐Ÿคฒ #securitydays #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใƒ‡ใ‚คใ‚บ #ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ
ใ€Security Days ๆฑไบฌใ€‘3ๆ—ฅ็›ฎใŒ้–‹ๅน•ใ—ใพใ—ใŸใ€‚ ไผšๅ ดใงใฏๆœ€ๆ–ฐใฎใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃๆƒ…ๅ ฑใ‚’ๆŽ˜ใ‚Šไธ‹ใ’ใ‚‹ใ€ๅ„ใ‚ปใƒƒใ‚ทใƒงใƒณใŒ้€ฒ่กŒไธญใงใ™ใ€‚ ๅคšใใฎ็š†ๆง˜ใซใ”ๆฅๅ ดใ€ใ”ๅ‚ๅŠ ใ„ใŸใ ใ„ใฆใŠใ‚Šใพใ™ใ€‚ ๅ…ฌๅผใ‚ตใ‚คใƒˆ๏ผšf2ff.jp/event/secd #securitydays #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใƒ‡ใ‚คใ‚บ #ๅฑ•็คบไผš #ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ
1
5
2
41
2,270
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
ๆœฌๆ—ฅใฎใ‚ชใƒผใƒ—ใƒ‹ใƒณใ‚ฐใ‚ปใƒƒใ‚ทใƒงใƒณใงใฏใ€ OpenAI Japan ๅฎ‡้ƒฝๅฎฎๆง˜ใซใ”็™ปๅฃ‡ใ„ใŸใ ใใพใ—ใŸโœจ ็ซ‹ใก่ฆ‹ใŒๅ‡บใ‚‹ใปใฉใฎ็››ๆณใถใ‚Šใงใ—ใŸ๐Ÿ˜‡ ใ‚ใ‚ŠใŒใจใ†ใ”ใ–ใ„ใพใ—ใŸ๏ผ
Security Days ๆฑไบฌใซใฆ OpenAIใ‹ใ‚‰ใ€Œใƒ•ใƒญใƒณใƒ†ใ‚ฃใ‚ขAIๆ™‚ไปฃใฎใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใ€ใซใคใ„ใฆ็™ปๅฃ‡ใ—ใพใ—ใŸ ๐Ÿ‘ฉโ€๐Ÿ’ป OpenAI ใฎ Codex Security Cloud ใงใฏใ€ใ“ใ‚Œใพใง็”ณ่ซ‹ใŒๅฟ…่ฆใ ใฃใŸ Daybreak Blue ใ‚‚ไฝฟใˆใพใ™ใ€‚ใ‚ณใƒผใƒ‰ใ‚’ๆ›ธใใ“ใจใชใใ€Cloud ๅ†…ใง AI ใซใ‚ˆใ‚‹่„†ๅผฑๆ€งใฎ็™บ่ฆ‹ใƒปๆคœ่จผใƒปไฟฎๆญฃใ‚’้€ฒใ‚ใ‚‰ใ‚Œใพใ™ใ€‚ ๆœ€ๅ…ˆ็ซฏใฎAIใ‚’ไปŠใ™ใใซๆดป็”จใ—ใ€็พๅ ดใฎ้˜ฒๅพกๅŠ›ใธ ๐Ÿคฒ #securitydays #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใƒ‡ใ‚คใ‚บ #ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ
2
3
417
Security Days ๆฑไบฌใซใฆ OpenAIใ‹ใ‚‰ใ€Œใƒ•ใƒญใƒณใƒ†ใ‚ฃใ‚ขAIๆ™‚ไปฃใฎใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใ€ใซใคใ„ใฆ็™ปๅฃ‡ใ—ใพใ—ใŸ ๐Ÿ‘ฉโ€๐Ÿ’ป OpenAI ใฎ Codex Security Cloud ใงใฏใ€ใ“ใ‚Œใพใง็”ณ่ซ‹ใŒๅฟ…่ฆใ ใฃใŸ Daybreak Blue ใ‚‚ไฝฟใˆใพใ™ใ€‚ใ‚ณใƒผใƒ‰ใ‚’ๆ›ธใใ“ใจใชใใ€Cloud ๅ†…ใง AI ใซใ‚ˆใ‚‹่„†ๅผฑๆ€งใฎ็™บ่ฆ‹ใƒปๆคœ่จผใƒปไฟฎๆญฃใ‚’้€ฒใ‚ใ‚‰ใ‚Œใพใ™ใ€‚ ๆœ€ๅ…ˆ็ซฏใฎAIใ‚’ไปŠใ™ใใซๆดป็”จใ—ใ€็พๅ ดใฎ้˜ฒๅพกๅŠ›ใธ ๐Ÿคฒ #securitydays #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใƒ‡ใ‚คใ‚บ #ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ
ใ€Security Days ๆฑไบฌใ€‘3ๆ—ฅ็›ฎใŒ้–‹ๅน•ใ—ใพใ—ใŸใ€‚ ไผšๅ ดใงใฏๆœ€ๆ–ฐใฎใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃๆƒ…ๅ ฑใ‚’ๆŽ˜ใ‚Šไธ‹ใ’ใ‚‹ใ€ๅ„ใ‚ปใƒƒใ‚ทใƒงใƒณใŒ้€ฒ่กŒไธญใงใ™ใ€‚ ๅคšใใฎ็š†ๆง˜ใซใ”ๆฅๅ ดใ€ใ”ๅ‚ๅŠ ใ„ใŸใ ใ„ใฆใŠใ‚Šใพใ™ใ€‚ ๅ…ฌๅผใ‚ตใ‚คใƒˆ๏ผšf2ff.jp/event/secd #securitydays #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃใƒ‡ใ‚คใ‚บ #ๅฑ•็คบไผš #ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ
1
5
2
41
2,270
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
OpenAI models are coming to Google Cloud. We're bringing our frontier models and Codex agentic workflows to Google Cloud's Model Garden, so enterprises can use our technology where they already build. openai.com/index/openai-modeโ€ฆ
487
410
232
5,267
569,895
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
OpenAIใฎใ‚ขใƒƒใƒ—ใƒ‡ใƒผใƒˆใซใคใ„ใฆใ€ใƒชใƒชใƒผใ‚นใƒŽใƒผใƒˆใงไธ€ๅ…ƒ็š„ใซ้–ฒ่ฆงๅ‡บๆฅใ‚‹ใ‚ˆใ†ใซใชใ‚Šใพใ—ใŸ๏ผไปŠใพใง็„กใ‹ใฃใŸใฎใŒไธๆ€่ญฐใงใ™ใŒใ€ใ“ใฎใƒšใƒผใ‚ธใ‚‚ใ—ใใฏRSSใ‚’ใƒ•ใ‚ฃใƒผใƒ‰ใ™ใ‚Œใฐๆœ€ๆ–ฐๆƒ…ๅ ฑใ‚’ใ‚ฆใ‚ฉใƒƒใƒใงใใพใ™๏ผopenai.com/ja-JP/products/reโ€ฆ
24
2
114
7,577
๐•ค๐•™๐• ๐•œ๐•  ๐•ฆ๐•ฅ๐•ค๐•ฆ๐•Ÿ๐• ๐•ž๐•š๐•ช๐•’ retweeted
OpenAI models helped researchersโ€”operating as part of our Patch the Planet initiativeโ€”discover critical security flaws in PHP, an open-source programming language used to run an estimated 70% of websites. Over the past two months, we led a coalition of maintainers, researchers, and security providers to prepare fixes and protections ahead of disclosure. This duration was necessary because the fix was quite intricate and required a major change to the PHP engine. Every site using PHP can now safely upgrade their version to stay secure, in addition to applying mitigations from various security vendors. Due to the severity of these flaws and PHPโ€™s widespread use, we consider this bug to be one of the most impactful vulnerabilities in recent memory. It also marks a new phase, where AI can find security bugs with internet-scale impact. This is an example of the Defenderโ€™s Window in action: to our knowledge, frontier AI was able to find this vulnerability ahead of attackers, giving defenders time to prepare. openai.com/index/coordinatedโ€ฆ
624
2,861
1,528
14,417
6,007,015