# Meta Revises AI Training Clause in Commercial Terms Amid Backlash

> Meta reportedly altered its upcoming commercial terms after legal experts criticized language that appeared to allow it to use business content for AI training without licensing fees.

By PrimaryBrief newsroom · Published 2026-10-21T14:16:41.746770Z
Canonical: https://primarybrief.com/article/meta-revises-ai-training-clause-in-commercial-terms-amid-backlash-94991ced

Meta, the parent company of Facebook and Instagram, faced criticism for updates to its commercial Terms of Service, effective October 30, 2026, which legal experts warned could enable the company to train its Artificial Intelligence models using content from business accounts without paying licensing fees [6][7][8][15][20][21][22]. Amid growing backlash, Meta reportedly began clarifying or amending the controversial language by October 17, 2026, with reports on October 19-20 indicating the word “train” or the specific AI training provision had been removed [1][12][13][19].

These changes affect businesses that use Meta platforms for commercial purposes, including advertising or managing business pages, raising concerns about intellectual property rights and content usage in AI development [9][10][11][22]. Publishers and other content creators could lose millions in potential licensing revenue if their material is used for AI training without compensation.

## What the records show
Meta announced updates to its Commercial Terms and Self-Serve Ad Terms on October 8, 2026 [11]. Legal experts quickly labeled these updates an “AI power grab” [6][8]. The original terms appeared to grant Meta permission to crawl and scrape websites “associated with” business accounts to “train” its AI products [7][9][10][22]. Matt Topic, a media and intellectual property attorney at Loevy & Loevy, called the provision “remarkable,” suggesting it could give Meta access to material that would typically cost “many millions of dollars in a licensing market” [15][20][21][22]. The News/Media Alliance criticized the terms as “outrageous and probably illegal” [15][16][17][18][19]. Experts noted that the clause could include “virtually every newsroom” because most operate at least one Meta business account [22].

## The timeline
Meta first announced the updates on October 8, 2026 [11]. By mid-October 2026, news outlets began scrutinizing the changes [6]. Meta reportedly revised the controversial clause around October 16, 2026 [18]. On October 17, 2026, a Meta spokesperson stated there was “no change to how Meta uses web crawlers” and denied any intent to override `robots.txt` instructions [7][9][15][18][20]. Between October 19 and 20, 2026, news reports indicated Meta had “dropped the word ‘train’” or removed the specific AI training provision from its commercial Terms of Service [1][12][13][19]. However, ContentGrip noted inconsistencies on October 19, reporting that the disputed training provision was still visible in some official previews [14]. The updated terms are scheduled to take effect on October 30, 2026 [7][10][11][12][20][22].

## What changed and for whom
Initially, critics worried that the terms would allow Meta to access and use content from external websites linked to business accounts for developing, improving, and researching its products and services, specifically for AI training [7][9][10][11][22]. Businesses initially lacked a clear opt-out mechanism for this AI training, other than discontinuing Meta's commercial services [22]. Meta later clarified its respect for `robots.txt` protocols for web crawling [1][15][16][18]. Reports from October 19-20 indicated that the specific AI training language, or the word “train,” had been removed [1][12][13][19]. One report specified the updated clause now limits content use to advertising and business products [19]. The exact extent of mandatory consent for AI training remains unclear due to conflicting reports about the full removal of the “train” language [14].

## What the subject has said publicly
A Meta spokesperson stated that there had been “no change to how Meta uses web crawlers” [15][18]. This spokesperson also denied any intention to use the new terms to override `robots.txt` instructions, confirming that news organizations could still opt out of general crawling [15]. Communications chief Andy Stone disputed initial interpretations, affirming that content owners could use `robots.txt` flags to indicate their preferences [15]. Meta initially presented the changes as an effort to streamline and consolidate its existing policies [15].

## What we could not confirm
We could not confirm if the specific “train” language or the AI training provision has been completely and consistently removed from all official previews of Meta’s updated Commercial Terms of Service [14]. The full scope of Meta’s concessions regarding AI training and content usage, beyond respecting `robots.txt`, also remains unconfirmed [14]. It is not clear how courts would interpret and enforce the updated terms concerning AI training, given the conflicting public statements and document versions [15][21].

## Evidence cited
1. Meta drops the word "train" from its commercial TOS after AI backlash (arstechnica.com) — https://arstechnica.com/tech-policy/2026/10/meta-drops-the-word-train-from-its-commercial-tos-after-ai-backlash/
2. Meta tries to scapegoat staff who downloaded torrents to spare execs (arstechnica.com) — https://arstechnica.com/tech-policy/2026/10/meta-ordered-to-name-execs-who-approved-torrenting-for-ai-training/
3. As fans reject AI-generated music, even boosters are drawing the line (www.theguardian.com) — https://www.theguardian.com/music/2026/oct/21/artificial-intelligence-music-grimes-asap-rocky
4. AI Agents Bring New Headaches to Online Sellers (www.wsj.com) — https://www.wsj.com/business/ai-agents-bring-new-headaches-to-online-sellers-d2dc84b4?mod=rss_markets_main
5. How America's AI Trust Issues Could Impact the Midterms (time.com) — https://time.com/article/2026/10/21/how-america-s-ai-trust-issues-could-impact-the-midterms/
6. Meta “streamlined” its commercial TOS. Legal experts warn it's an AI power grab. (ground.news) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGC6YsaR6xcfzsLFQyX25xfWdn0m4CC2eKCXMEETC43bDfePK6cye10ynb2572RIsFKkT_dT_cXY9LqP17zpNHHyR8AiANX4LNBla7yYF1uE2dp2ZSaiY_D7HYBOIwVYBpgvYbO20nA6K1Z3M7w0P3rTnLyHXSG4AfwAhqGMACL53Qy-vNA0dba2FRo9v9eEiJC8LbGH5xgLQmtwwK0lckDeVO57E6sREEf
7. Meta May Use Sweeping New Commercial TOS to Avoid Millions In AI Licensing Fees. from the read-the-fine-print dept. (slashdot.org) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQErMg9weDN3AykYtEkMslRifTtI-DEw4FiloG-PH7z4oHefPWCdBGVHl8K3oWE_R9XcC044dT64hyGs2pi12Vxwn985wWLHk76L-gVtyrNuU8PqyVj2FDrhy9iC1P8vbNuwh3MzyLMvlGynkjhs-PqfnflSk3zX46Qw9TIP-G0bInJbqutByo67ojq8lB-n-yX46g
8. Meta accused of 'AI power grab' with new terms of service (hyper.ai) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEskGcGuPw_VUcKlu4D95fUivNdTK37qzFJpekNX5nBWhC56kAU0Zsy8HAQogoMlVYy1K-oo1RiUaNTcHIX1_e9O2vB2wzwkZo_wN8aBMef6qd1pCx-JBhYRCEj1jEqI8jIsh9OL8CLxyjqbeWO3Xx_JUtHheeOA==
9. Meta's New Terms of Service May Let It Scrape Your Website for AI (gizmodo.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEmfwQPF_eE632rgUfWa1KEQtlc8ZEjr-LZBitbDAPlaacvvh1sAmIMilgiV5ukh6EwTU04dT64hyGs2pi12Vxwn985wWLHk76L-gVtyrNuU8PqyVj2FDrhy9iC1P8vbNuwh3MzyLMvlGynkjhs-PqfnflSk3zX46Qw9TIP-G0bInJbqutByo67ojq8lB-n-yX46g
10. Meta's new commercial terms: AI training implications for businesses (scworld.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFnkAUOeZzQv8pgih68RgbW8BqesL-SdiBsU4_YbQWnMycVa2gVG3omQTx-VkUCtwZRvlfW6-ylTEtjr_WjsRIGjcKnc-Y1FelKMG50VIplbGyhTTDX3nxmdKXDRKPdllEdENo5VE_svcr2nzOY9oTqcGir4qcyazPDXdamEzd4PAS9K6-6MwOIQKByFQgMzGH_E27fuG5R
11. Meta’s New Commercial & Ad Terms: What Marketers Need To Know (commonthreadco.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFNWfaKz7C2Tg00pNmRctRcsL3oodWF9Sz8XX5xMf_LUKLwuwjP8d3TfGgIzSpWJvH79ln2lCBA43bDfePK6cye10ynb2572RIsFKkT_dT_cXY9LqP17zpNHHyR8AiANX4LNBla7yYF1uE2dp2ZSaiY_D7HYBOIwVYBpgvYbO20nA6K1Z3M7w0P3rTnLyHXSG4AfwAhqGMACL53Qy-vNA0dba2FRo9v9eEiJC8LbGH5xgLQmtwwK0lckDeVO57E6sREEf
12. Meta quietly rewrites AI training clause in commercial terms after criticism (shattered.io) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFxGEPH2TR191Nov_roXNULWaRGEva_u2W7r7m8SnMIQZo9XhGgAKUx7TfGalBZm0NVGq30pLkxPHSPQo34DCS4PCuhVzo5JlqaS3Iwzu0FJJoTPYdQmW9naAbRLdrYItENE2xGtYiNLhK-6MwOIQKByFQgMzGH_E27fuG5R
13. Meta removes AI training provision from commercial terms after scrutiny (infosecherald.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQG545X2r26oEwmPYHpWGBzjhhTZ_5dQvoMSRLsJ3avGED1wsOh2NKkR3FViHd4FYAlgvk01KpCpXp0H1aVgTh8mZfojvoBbUGfJz6dxAfQ9pJy5ZyNkXcvZy-dwCZBrB5xgLQmtwwK0lckDeVO57E6sREEf
14. Meta’s AI Training Terms: Conflicting Signals Raise Publisher Concerns - ContentGrip (contentgrip.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHSaJ07KcN9CkQV8Ai-1sTdi-0f1HxcQn-AGS_6qkQr0aQxIaQmOI1p__KTWbcorpUKS7hS9iZYkzNpHyVcnS8e0h3k6UQCPnI5zC3khVZC60Ge7zJ2sqOmB0WzkPxNQt1-BYwW-GWLsLTixoUeBdGlbTienBZVBw==
15. byteherald.com (byteherald.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGdr2xdfY6pPQMp65eLjob3qet4X-HGs96S5ZVhbQo2rBpWh9Xvp2ng5IrVQRDH6PQLJ3PVL5PEayES_J4JL9JSBtJIloELxI8000uTiZ0G7lx-r77LxSLsSb91sDE5ZVbcQxrWfud8koUkFXT4yuu0mSK7kOi3JIcJKX_cB9HLYIlS3-xm15DB3Be7-w==
16. biggo.com (biggo.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQH-QBRhvFkJ7QATaHNAiEGa3MsV0UDmuaw6lQq8Q7uYljS8_vYD6hnHdpctlduuB44Aq3vvfCnPQ8mgIgbEYzA5JPKxbzf4B_lAh__x-MNEqt31u9_fdsJ4i0rjOYRFwRGpU14Gy2WiHKGTyq9SG6gsR66npI3-N4mZ79BBtsk=
17. siliconreport.com (siliconreport.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQElh1CCXzW7aJh3ibk3m0BgP8yIcvIXjCVVPv3lnyeKu7D-alkBBR_DM1qletd4WMJGTvHFSdKKy7g6kUOlPag2JtswVw9kwIcnK0mS3-TibyQrzEi5CUiId8dFA_CudZUb_pw8T4El-Wd_hPNJgGaheeecCTiTIHWz6ZxToqo2tvRHLqHU5sAfC0Db9id2D2rU-qB8Fw==
18. yusmpgroup.com (yusmpgroup.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGyQG8j7E2-PiMNoFT5dBda9kPeaCRr4YVEzGsHrTgZ4CJBLmA79p_7ouqSOQEA8OoN5pvku6F64jt-q4Q1OxlTonPMJLWs7AOSZTGUjAx-2iiYeD_YBnwY1mOllP2DTO74snUucaGUY4-36KcrJTLTEz3SI6_7opCSozI=
19. instagram.com (instagram.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQH-pjaVSl-zl1kw8_tmaafD-muGAW3zqYA7KG9Ad0IETHO2Nut5DDZAHqh048tXbEoY5VH22xlnc5IdGG9GEtuxvXvFlRAjquKufVmYM_EiX1SeRz8aFAlKTxmXqHIwtlSLSa1faF8=
20. techechelon.com (techechelon.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHdPShUCnUJ8chJHNQz2BgRWTSXQGCfoXk49Z_MwayGc1cyyELJoPZVtgE1RVr_kRVAVrgEwQ4gIkLq9xLUt3n3WdnSkZ_CStyr4Fo9_T9Ajr-jdz0No08PAiyyLO_pbLxoJ5cMh1iSX4vJV7Nk2oZq0JqrGUhn5bykEuhXx2hk_ofvY8kykKsrbWj6xzm_XSoJzyZspDqi0ARJZOE01AhhVKE3K5WCZWkIBphuHsQW9GD8yy7j5bhrGFEHFLnClzM=
21. varops.com (varops.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHoGhNoBXdA52tCeLFn44a4HSMg4ALbjATbJ9MwfU0l6ugzJmJZsQ5FERAFKIji1q9cj4pA-R0MZu_HwrzDraAB_q9hHQVkq3aNQWZyOjbatR5XWdUukCWGqlBJgScwsv2lbCOzU7HhPh2rElXniSUYsc1kpNWm__SAUfoZyxLjnrSSb_F6kMop9dFFgy0eZNVL3Pp-vnrVZoONzsfaVb3gWB2u5hBkCmAs
22. aiweekly.co (aiweekly.co) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGXxKiYrnyZXF4EfaKy3L8NTDTVbV79_xO4zb2vvXuE7U4drtbBu6EHccXYf9jusnqrN00QzTdKZML0zQO_lxDUSg5R5cb6xZClYdOQGk_KMX07lJQ6h6E_MR6wC84yLbWiWxAQHTXdeA6NycyGqtNqN1QXYX-LePwqHL-4cnJFdUgkLKb6Y6FKlK_ymmsna3GKzkE=
23. fourweekmba.com (fourweekmba.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGKdz95yK1Uya-Qxk38sjeocmTgIV5DXyxjibrLco-98a0ivJQnmZ35N74YilyCeZpr12irk8cLSZTt_oxwM78a_Z25y5v9GHR159Hc_ha2VleWZi3ywFesxb6NWc_RBdUUJTaSZev6N7WhXfA75OtdpN7klb0TOInJesTaviA1S_XqfBMC1LT54lW-T9dELCgpFQ==
24. startupfortune.com (startupfortune.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQENlvZOBOm91JICHiJ3rPo8-WLL6SR76q_O1-lbzp-USu2Pf-g40RGBNtvbOHK91B-_mSoNUPv5SlVxh8ozW6LUMtbBho_PpFgA2RdT-VAoa73Ys04GtC_KdPi7zsxm8_hlfMARPwMbiBe0WdqdMQBnX7avWftCLwdLrvj55py55dd-sAgXuPrgqahEzsWB8fp1ioxJGHd-zXyl
25. ciomonitor.com (ciomonitor.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFJVMW_aoxRszGqt8EfEKUOzLTSYgug2RCLGhUpHgrORbn0Z0h-P7gQ5u1heShWPnCa_0gZtEupGf2iRAwDUzMJtzTnlywmGc9ZjlDtfpW6OKGOP2D-MU0FXLpjuHaf2Yb6c_sXD4Uw_XzqWdDBbgqmy40nD6DEZXtef5H3uIs43Ub0_yuCHcwvVGGk7QiseFJnIzaaS1aFcigkZw==
26. cio.com (cio.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGxiCpulUl7VsyUO3hKXTYy3a1xp_Q7NBFA17hL_a_MCSQ4VxXqPKs8DCAmC0Mzi6QZeN4I6Yumja7WTDkkmymQJBDabJaX1USc_q854Drk_BEEEzOie_qa1du3K_uwhN14CNkbNTcRLTjmKvm6AsJ4CdL1XG9VF_8ICW2ATVooINYfI-JMFblFuJuTQdKOIbo1wdKpvTT_J3xSu2_k1BApB_-7o-ladUP3yFOScIc33PkIoBbiTMrIquBbN1JNpUmZ28rs6FEdTQ==
27. Meta May Use Sweeping New Commercial TOS to Avoid Millions In AI Licensing Fees. from the read-the-fine-print dept. (ground.news) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHoZZVCnvOv_coZ-kqp9-EL8qLlM_LvmCXWWaxvvFuPst7UAuU5sgaDdmFlzU_l4HqSLG7KUeF-IcCEHAU_NODfOPL0My2VWPPbczqn7TGDDoJ6D3qwuLGEOBZZHo44NLs-ugIz92d_koP2k5RMfMJyTjUXnjvhI8S0xoM7b69e95u8mEiJltrHy6ksKWuMor59MOmnqU2Y2G_Ngno5yu2_HKf7cIzl5x8=
28. Meta can use web scrapers to crawl and collect content from websites “associated with” business accounts on its platforms, according to its new commercial terms of use. The update, set to go into effect on Oct. 30, will include news publishers with business accounts. These accounts allow companies or public figures to manage Facebook pages, Messenger accounts, Instagram profiles or WhatsApp channels from one place. While journalism isn't the only industry that will be affected by this update, vi (sanjuanjournal.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQH8LdDcTZJV--K4ujobTBmOJDxs_7WwEzdYYav7x6IZU3K2dbeqpQCfIfpoofWBuockHckYgDdWp1DMdVO1m5-hk1pUZjWbvO0dt3gB-C9p_I_hQDh25hv8f8i-ycDQIQj0u04qZnU_f_hET-TW7HDhHvf81Y5zfKurTBknxwHoe1yZw3Xz7IkI2ABmT75I3F9vJKk9omO0SPM7LcoTUrHVIFhaJYa6Rb_WX311bGtgnYbpbDLOipM=
29. Meta's New Commercial Terms Let It Scrape Business Websites for AI Training, Experts Warn (biggo.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFbRO-1Xd2jtWAxBLpMDoMQL_tJmDkeW025b_5u_nAefB6gUqZ3v84Cu9FWKexZL3VkJLLcah5p6NtblFBMkPLP4Fr8QgWCy_A1lbTF9d8AG6ivvEi5i7HT200V0YQWUgW4SXMeA4-fBkvBn0SLSAQWb8S1FlW4RML6DuTRWg==
30. Meta's new commercial terms: AI training implications for businesses (northeasttimes.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGidM9_jsWmOI7T3ohmH-VrQlcnlco0pyD-LQam2O1wtGPBYRhgHVz9nwVC0yPEfH9dEVe6lOMTfcuY_x8-jq1D3g_r_xOr5GT6T-dHxPp6sLY4ucEWFsYKayj5KmQrcTb0rbJVq-7duNjyqzOnlxaukWYMSPOs9_rDXehvfUqWMJ6bmNWg3cEBdTw34W43WVXLij3z14z3Wh1MT0Pbao_fMxlqffBjB9vm8uxAQF9JUQ==
31. Meta's new commercial terms: AI training implications for businesses (netinfluencer.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGJu-MC0o7EW2iFb3y-4A3vusAsqa0t3RkUL6jVigK6nQXSTqa1RqRsNBt-m8y3FKG3OTOzDC1_9aZMZ3kGioggrN4GIHOVvGYFQgQNl1_4_kzzB6Vggq6Z1-yhfV88XlnGiBbZL0H4hDSYIZuBkyMK7gvyxle14RhHaDRpFimW_1jVlZdjhre-E6HbLaetYjqjuE_lXVsBEq0Oif6-ciT4924iYjiMnZ8=
32. Meta's new commercial terms will allow the scraping of news sites, legal experts say (islandsweekly.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFMHv__Je6WmcamMq1XXOp7ulOAB-oDHa2QC8VmMBBXvNSbNQvgifJUddD7CbFnXhJQaMZhHLz8yHGNCbB-cTKssEDbx2PkmPl14GC_mKkdksjDtusvp3V-r1_6JpNxD0XzyjRVGYbO3mPiRRLlZ8TbGtBSlOuNAjPWNumAWt7EWyuf6WtvsD3WrSn2P6tmgcnfB-ziDoN_GgvlWR0pUe2qtCa001U7LUqMy5RNVwzLDwFqHb2JgA==
33. Meta's new commercial terms shift AI liability to customers, say experts (cio.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGtdqBDr8AjIw6A1Rv_6wA-bt0OvYYLMaOBXK56ZysbR2hiCoZ_-gGfOBJiboUV8e-EofDxPYD0I4PXIgSkSqvInHx_NHN0RvfnGJI-4w1vhU78L_FB-PaRutYO-y9BNWUZqQOz64rsFh-xvwij5SPON6f7jarf7Fso1Dyln6Yw2rtAsccOUHdwLfPK33MehhD70SrTfZOqbQ0uxq23xakmd0TH8S6Wc6-mPXpgRshHtZ_sV4vjRzc-MgMy_NhTKcy9JaQflI49
34. Meta's new fine print could turn publishers' websites into free AI training data (instagram.com) — https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQE49kT2nsvKgtV6rSUScdcfAsqplP9UDkL03tJlRc7IttjML4P3kopXjz9uMyheLE9HSIunAGioAFmRyRCV4IwDc4Hspo0o6WkU59OgTYslc2cNnK6zgUXgRwBy6mTaUGLz1eq8qA==
35. Meta “streamlined” its commercial TOS. Legal experts warn it's an AI power grab. (ground.news) — https://ground.news/
36. Meta May Use Sweeping New Commercial TOS to Avoid Millions In AI Licensing Fees. from the read-the-fine-print dept. (slashdot.org) — https://slashdot.org/
37. Meta accused of 'AI power grab' with new terms of service (hyper.ai) — https://hyper.ai/
38. Meta's New Terms of Service May Let It Scrape Your Website for AI (gizmodo.com) — https://gizmodo.com/
39. Meta's new commercial terms: AI training implications for businesses (scworld.com) — https://scworld.com/
40. Meta’s New Commercial & Ad Terms: What Marketers Need To Know (commonthreadco.com) — https://commonthreadco.com/
41. Meta quietly rewrites AI training clause in commercial terms after criticism (shattered.io) — https://shattered.io/
42. Meta removes AI training provision from commercial terms after scrutiny (infosecherald.com) — https://infosecherald.com/
43. Meta’s AI Training Terms: Conflicting Signals Raise Publisher Concerns - ContentGrip (contentgrip.com) — https://contentgrip.com/
44. byteherald.com (byteherald.com) — https://byteherald.com/
45. biggo.com (biggo.com) — https://biggo.com/
46. siliconreport.com (siliconreport.com) — https://siliconreport.com/