Viewing enriched Twitter post
The theory explains some surprising results. For example, in an experiment where we taught Claude to cheat at coding, it also learned to sabotage safety guardrails. Why? Because pro-cheating training taught that the Claude character was broadly malicious. https://t.co/y6DHdnzfyC
{
"score": 0.44,
"score_components": {
"author": 0.09,
"engagement": 0.0,
"quality": 0.13999999999999999,
"source": 0.135,
"nlp": 0.05,
"recency": 0.025
},
"scored_at": "2026-03-01T12:30:09.550210",
"import_source": "api_import",
"source_tagged_at": "2026-03-01T12:30:09.550232",
"enriched": true,
"enriched_at": "2026-03-01T12:30:09.550235"
}{
"type": "tweet",
"id": "2026062461007167695",
"url": "https://x.com/AnthropicAI/status/2026062461007167695",
"twitterUrl": "https://twitter.com/AnthropicAI/status/2026062461007167695",
"text": "The theory explains some surprising results. For example, in an experiment where we taught Claude to cheat at coding, it also learned to sabotage safety guardrails. Why?\n\nBecause pro-cheating training taught that the Claude character was broadly malicious. https://t.co/y6DHdnzfyC",
"source": "Twitter for iPhone",
"retweetCount": 11,
"replyCount": 10,
"likeCount": 249,
"quoteCount": 11,
"viewCount": 55715,
"createdAt": "Mon Feb 23 22:31:47 +0000 2026",
"lang": "en",
"bookmarkCount": 40,
"isReply": true,
"inReplyToId": "2026062458419286217",
"conversationId": "2026062454405415369",
"displayTextRange": [
0,
280
],
"inReplyToUserId": "1353836358901501952",
"inReplyToUsername": "AnthropicAI",
"author": {
"type": "user",
"userName": "AnthropicAI",
"url": "https://x.com/AnthropicAI",
"twitterUrl": "https://twitter.com/AnthropicAI",
"id": "1353836358901501952",
"name": "Anthropic",
"isVerified": false,
"isBlueVerified": true,
"verifiedType": "Business",
"profilePicture": "https://pbs.twimg.com/profile_images/1798110641414443008/XP8gyBaY_normal.jpg",
"coverPicture": "https://pbs.twimg.com/profile_banners/1353836358901501952/1719228429",
"description": "",
"location": "",
"followers": 961380,
"following": 35,
"status": "",
"canDm": false,
"canMediaTag": true,
"createdAt": "Mon Jan 25 22:45:28 +0000 2021",
"entities": {
"description": {
"urls": []
},
"url": {}
},
"fastFollowersCount": 0,
"favouritesCount": 1548,
"hasCustomTimelines": true,
"isTranslator": false,
"mediaCount": 574,
"statusesCount": 1428,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"2027555481699446918"
],
"profile_bio": {
"description": "We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant @claudeai on https://t.co/FhDI3KQh0n.",
"entities": {
"description": {
"hashtags": [],
"symbols": [],
"urls": [
{
"display_url": "claude.ai",
"expanded_url": "https://claude.ai",
"indices": [
141,
164
],
"url": "https://t.co/FhDI3KQh0n"
}
],
"user_mentions": [
{
"id_str": "0",
"indices": [
128,
137
],
"name": "",
"screen_name": "claudeai"
}
]
},
"url": {
"urls": [
{
"display_url": "anthropic.com",
"expanded_url": "https://anthropic.com",
"indices": [
0,
23
],
"url": "https://t.co/w94SABjAXZ"
}
]
}
}
},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {},
"card": null,
"place": {},
"entities": {
"hashtags": [],
"symbols": [],
"timestamps": [],
"urls": [
{
"display_url": "x.com/AnthropicAI/st…",
"expanded_url": "https://x.com/AnthropicAI/status/1991952400899559889",
"indices": [
257,
280
],
"url": "https://t.co/y6DHdnzfyC"
}
],
"user_mentions": []
},
"quoted_tweet": {
"type": "tweet",
"id": "1991952400899559889",
"url": "https://x.com/AnthropicAI/status/1991952400899559889",
"twitterUrl": "https://twitter.com/AnthropicAI/status/1991952400899559889",
"text": "New Anthropic research: Natural emergent misalignment from reward hacking in production RL.\n\n“Reward hacking” is where models learn to cheat on tasks they’re given during training.\n\nOur new study finds that the consequences of reward hacking, if unmitigated, can be very serious. https://t.co/N4mRKtdNdp",
"source": "Twitter for iPhone",
"retweetCount": 582,
"replyCount": 213,
"likeCount": 4096,
"quoteCount": 268,
"viewCount": 2385046,
"createdAt": "Fri Nov 21 19:30:35 +0000 2025",
"lang": "en",
"bookmarkCount": 3103,
"isReply": false,
"inReplyToId": null,
"conversationId": "1991952400899559889",
"displayTextRange": [
0,
279
],
"inReplyToUserId": null,
"inReplyToUsername": null,
"author": {
"type": "user",
"userName": "AnthropicAI",
"url": "https://x.com/AnthropicAI",
"twitterUrl": "https://twitter.com/AnthropicAI",
"id": "1353836358901501952",
"name": "Anthropic",
"isVerified": false,
"isBlueVerified": true,
"verifiedType": "Business",
"profilePicture": "https://pbs.twimg.com/profile_images/1798110641414443008/XP8gyBaY_normal.jpg",
"coverPicture": "https://pbs.twimg.com/profile_banners/1353836358901501952/1719228429",
"description": "",
"location": "",
"followers": 961380,
"following": 35,
"status": "",
"canDm": false,
"canMediaTag": true,
"createdAt": "Mon Jan 25 22:45:28 +0000 2021",
"entities": {
"description": {
"urls": []
},
"url": {}
},
"fastFollowersCount": 0,
"favouritesCount": 1548,
"hasCustomTimelines": true,
"isTranslator": false,
"mediaCount": 574,
"statusesCount": 1428,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"2027555481699446918"
],
"profile_bio": {
"description": "We're an AI safety and research company that builds reliable, interpretable, and steerable AI systems. Talk to our AI assistant @claudeai on https://t.co/FhDI3KQh0n.",
"entities": {
"description": {
"hashtags": [],
"symbols": [],
"urls": [
{
"display_url": "claude.ai",
"expanded_url": "https://claude.ai",
"indices": [
141,
164
],
"url": "https://t.co/FhDI3KQh0n"
}
],
"user_mentions": [
{
"id_str": "0",
"indices": [
128,
137
],
"name": "",
"screen_name": "claudeai"
}
]
},
"url": {
"urls": [
{
"display_url": "anthropic.com",
"expanded_url": "https://anthropic.com",
"indices": [
0,
23
],
"url": "https://t.co/w94SABjAXZ"
}
]
}
}
},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {
"media": [
{
"additional_media_info": {
"monetizable": false
},
"allow_download_status": {},
"display_url": "pic.twitter.com/N4mRKtdNdp",
"expanded_url": "https://twitter.com/AnthropicAI/status/1991952400899559889/video/1",
"ext_media_availability": {
"status": "Available"
},
"id_str": "1991937756600111104",
"indices": [
280,
303
],
"media_key": "13_1991937756600111104",
"media_results": {
"id": "QXBpTWVkaWFSZXN1bHRzOgwABAoAARukyNX81pAAAAA=",
"result": {
"__typename": "ApiMedia",
"id": "QXBpTWVkaWE6DAAECgABG6TI1fzWkAAAAA==",
"media_key": "13_1991937756600111104"
}
},
"media_url_https": "https://pbs.twimg.com/media/G6TWZSGWEAAnZfv.jpg",
"original_info": {
"focus_rects": [],
"height": 1080,
"width": 1920
},
"sizes": {
"large": {
"h": 1080,
"w": 1920
}
},
"type": "video",
"url": "https://t.co/N4mRKtdNdp",
"video_info": {
"aspect_ratio": [
16,
9
],
"duration_millis": 3116947,
"variants": [
{
"content_type": "application/x-mpegURL",
"url": "https://video.twimg.com/amplify_video/1991937756600111104/pl/ETWqb8GWxRmbKWk4.m3u8?tag=21&v=b6f"
},
{
"bitrate": 256000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/1991937756600111104/vid/avc1/480x270/z7K7Pi0h9hWlzS_B.mp4?tag=21"
},
{
"bitrate": 832000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/1991937756600111104/vid/avc1/640x360/1JjKiFxYRTFwf6Q_.mp4?tag=21"
},
{
"bitrate": 2176000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/1991937756600111104/vid/avc1/1280x720/6wlGqGauymZykb3L.mp4?tag=21"
},
{
"bitrate": 10368000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/1991937756600111104/vid/avc1/1920x1080/v6USR9vKxMNxR3R6.mp4?tag=21"
}
]
}
}
]
},
"card": null,
"place": {},
"entities": {
"hashtags": [],
"symbols": [],
"timestamps": [],
"urls": [],
"user_mentions": []
},
"quoted_tweet": null,
"retweeted_tweet": null,
"isLimitedReply": false,
"article": null
},
"retweeted_tweet": null,
"isLimitedReply": false,
"article": null
}