《硅谷银行：2024网络AI技术的兴起：生成式AI和网络安全的交汇报告（英文版）（19页）.pdf》由会员分享，可在线阅读，更多相关《硅谷银行：2024网络AI技术的兴起：生成式AI和网络安全的交汇报告（英文版）（19页）.pdf（19页珍藏版）》请在三个皮匠报告上搜索。

1、May 2024TITLE OF THE REPORT|23Letter from the Authors4Macro9Fundraising&Investment12Survey Results15ExitsAlmost exactly a year and a half after the public release of ChatGPT,the new frontier of Generative AI(GenAI)has captured public imagination.For business leaders,the increasing proliferation of G

2、enAI creates a distinct set of challenges and opportunities as companies begin exploring ways to augment or even replace existing workflows in the enterprise.While we have started to see GenAI affect everything from how companies develop software to how they engage with their customers,we are ultima

3、tely still only at the outset of what may end up being a platform shift.This shift could be as influential as the advent of mobile or cloud computing proved to be over the past two decades.How,though,does this emergence affect how Chief Information Security Officers(CISOs)protect their organizations

4、?How do security leaders think about this burgeoning platform shift amidst an ever-evolving threat landscape?To what extent is the ability to secure GenAI applications an impediment to their being more widely adopted?These are some of the questions that our clients and their customers continue to gr

5、apple with.Rather than focus solely on GenAI and its proliferation,we have opted instead to explore the intersection of GenAI and cybersecurity or,said another way,“CyberAI.”Nearly 70%of CISOs polled for this report identified GenAI adoption as at least a top-five priority to their organization.Mean

6、while,almost all of the 26%of polled CISOs who have not yet changed their cybersecurity stack in response to increased GenAI proliferation expect to within the next three years,implying ample opportunity for new startups to help fill that gap.These are just two tailwinds that we expect will drive mo

7、re attention to CyberAI and accelerate startup formation.We believe that continued urgency to employ GenAI solutions will be a defining trend in the coming years and that CyberAI will be increasingly critical to the emergence of this new landscape.3Andrew McCartyVice PresidentTech BankingSilicon Val

8、ley BankEmma EschweilerDirectorInvestor CoverageSilicon Valley BankTHE RISE OF CYBERAI|Natalie FrattoManaging DirectorInvestor CoverageSilicon Valley Bank*We expect that our conceptualization of GenAI will evolve as the space continues to emerge.If you would like to contribute further to this defini

9、tion,please reach out at .Tools and platforms that leverage GenAI to address vulnerabilities in AI infrastructure,models and applications,or to defend against external,GenAI-driven threats and attacks.CyberAInoun sahy-ber ey-ahy4THE RISE OF CYBERAI|5Notes:1)According to SlashNext,there was a 1,265%i

10、ncrease in phishing from Q4 22 to Q3 23.Sources:The New York Times“Microsoft to Invest$10 Billion in OpenAI,”IBM“IBM Acquires Manta Software Inc.,”Federal Bureau of Investigation“Internet Crime Report 2023,”SlashNexts“The State of Phishing 2023,”PitchBook Data,Inc.and SVB analysis.THE RISE OF CYBERA

11、I|Support for CyberAI InfrastructureGovernment funding from programs in the CHIPS and Science Act and the Department of Defense cybersecurity budget are major drivers of investment in the infrastructure necessary for growth in the CyberAI space.Investment from CorporatesLarge corporates have grown t

12、heir research arms over the past decade to support AI model growth.Microsoft has invested$10B in OpenAI,and IBM has acquired 30+companies since 2020 to bolster AI capabilities.Ubiquity of ChatGPTReleased in 2022,ChatGPT brought GenAI to the masses.While not the first large language model(LLM),its wi

13、de adoption marked a long-term tectonic shift in how businesses and consumers interact with the technology.Mounting Costs of CybercrimeCybercrime losses increased nearly nine-fold from 2017 through 2023,spiking in the post-COVID era.This has spurred increased cybersecurity demand and spend,driving i

14、nvestment in the space.Shifting Attack SurfacesMany GenAI apps were built without security in mind,increasing potential vulnerabilities.AI researchers continue to advance data security,but much of what underpins a working LLM remains vulnerable.Use of AI by Bad ActorsAI increases efficiency not only

15、 of legitimate enterprises,but also of cybercriminals.Today,scammers can generate believable phishing emails at scale.In 2023 alone,phishing emails increased 13-fold.1Formation of CyberAI StartupsNew companies are forming an emerging CyberAI ecosystem.Between 2020 and 2023,the number of CyberAI comp

16、anies formed was nearly 2x the number from the previous four-year period.VCs Investing in CyberAI CompaniesCapital is flowing to these new CyberAI companies.Venture capital(VC)investors continue to pump money into cybersecurity and AI startups,allowing them to maintain high burn and focus on growth.

17、Investor Interest in CyberAI FundsRecognizing growth in the sector,investors are strategically focusing on CyberAI.Among US VC funds raised in 2023,those mentioning AI as a specialty were nearly twice as likely to close their funds as those that didnt.In the early 2000s,every pitch deck had to give

18、a nod to the companys mobile strategy as smartphones became ubiquitous.Today,ChatGPT has ushered in a new era of AI.There are now more than 7,500 US VC-backed companies that have AI as part of their goods and services,and investors are increasingly incentivized to focus their attention on this space

19、.While not the first LLM,ChatGPT brought GenAI to the masses,marking a long-term tectonic shift in how businesses and consumers interact with the technology.Against this backdrop,companies are eagerly adopting AI to realize the potential automation and augmentation of human working hours.At the same

20、 time,deploying new AI technology is expensive,both in terms of implementation and security.This could be why the largest companies are most likely to use or plan to use AI in the near term.Cybersecurity is a primary concern among companies,as the aggregate losses from cybercrime have risen nearly n

21、ine-fold in the past six years alone.To combat this,companies are investing an increasing amount in cybersecurity,with the bulk going to security services.While the recent spike in AI investment may reflect a bit of bluster,AI has the potential to become a horizontal platform creating efficiency gai

22、ns across all industries.6Notes:1)Companies reporting use of AI in producing goods and services in the last two weeks(currently using AI)or that plan to use AI in the next six months.Sample includes US-based respondents to the Business Trends and Outlook Survey.Employer size buckets defined by the U

23、S Census Bureau.2)Estimates based on National Statistical Institutes and O*NET via Accenture Research.3)Complaints to the FBIs Internet Crime Complaint Center.Sources:US Census Bureau,PitchBook Data,Inc.,Accenture“Work,Workforce,Workers,”Gartner Forecasts on Global Security and Risk Management Spend

24、ing,Federal Bureau of Investigation“Internet Crime Report 2023”and SVB analysis.Number of ComplaintsAggregate LossesTHE RISE OF CYBERAI|$1.4B$2.7B$3.5B$4.2B$6.9B$10.3B$12.5B302,000352,000467,000792,000847,000801,000880,0002017201820192020202120222023$215B$188B$165B202420232022Security Services42%Inf

25、rastructure Protection16%Network Security 11%Identity Access 9%Consumer 4%Other19%Current Use of AI0%2%4%6%8%10%12%1-45-910-1920-4950-99 100-249250+AllPlanned Use of AI in 6 Months+1.5+1.1+1.6+1.7+2.1+3.0+3.8+1.5Largest Companies Most Likely to Use AICompany Size(Number of Employees)33%33%28%23%22%1

26、6%22%18%35%34%22%19%19%23%16%12%SoftwareBankingRetailHigh TechLife SciencesHealthAerospaceIndustrialsAutomationAugmentation%of All CompaniesWhat are investors really investing in when they invest in AI?Most of the companies in the AI universe are not developing the core technology,but rather finding

27、 ways to apply it.Part of this is due to an urgency among startups to present themselves as AI-enabled even if AI is not their core offering.Among the nearly 700 US companies involved in AI that have raised more than$50M,less than one-third are building LLMs or creating new computing infrastructure.

28、1 Still,leveraging AI especially at the enterprise level is a complex endeavor,and cybersecurity is stalling broader AI adoption.At the forefront of this risk is data security.Whether in the databases or code underlying AI applications,security can often be weak,exposing both developers and end user

29、s to risk.In fact,in a survey of 281 research studies on this topic,researchers identified several vulnerabilities in LLMs,ranging from security and privacy concerns to inherent model vulnerabilities such as data poisoning and backdoor attacks.AI researchers continue to make advancements in data sec

30、urity to address these weaknesses,but much of what underpins a working LLM remains vulnerable.Cybersecurity companies that address these issues have attracted more investment over the past decade.However,scaling can be difficult among early-stage cybersecurity firms,as corporates often purchase star

31、tups for their tech before the startups can fully scale and widely deploy their security products.7Notes:1)Aggregated VC investment and valuations for US companies that have raised more than$50M in VC with the latest deal since January 2021.Segments based on PitchBook Data,Inc.taxonomy.2)Technology

32、segments determined using PitchBook Data,Inc.s vertical taxonomy.Sources:Yifan Yao et al.“A Survey on Large Language Model Security and Privacy”(High-Confidence Computing,2024),PitchBook Data,Inc.and SVB analysis.THE RISE OF CYBERAI|$8B$67B$116B$20BCommercial ApplicationsCore Technology$166B$286B$41

33、9B$89BRobots and AVsAppsAI ModelsComputingAI SegmentVC RaisedAggregate ValuationMobileCleantech 1.0Web3AIDot com eCommerceThe arrival of ChatGPT supercharged AI investing.Post-ChatGPT,AI investing has become its own beast.0.470.84Post-GPTPre-GPTR-Squared:AI VC&All VCPre-ChatGPTPost-ChatGPTCybersecur

34、ity0%5%10%15%20%25%19951996199719981999200020012002200320042005200620072008200920102011201220132014201520162017201820192020202120222023Cyber investment has held steady as many cyber startups get scooped up by larger companies before they can scale.$0B$10B$20B$30B$40B$50B$60B$70B202320222021202020192

35、018CybersecurityAIAI Investment Growth:163%Cybersecurity Investment Growth:36%Scale:=$10B=$2.3B=$3.7B8Notes:1)For US VC-backed artificial intelligence/machine learning(AI/ML)and cybersecurity companies.Only companies with valuations over$250M are shown for AI/ML and cybersecurity companies.AI/ML and

36、 cybersecurity verticals and subsegments determined by PitchBook Data,Inc.taxonomy.Circles represent valuation scale within each category(AI/ML,cybersecurity and CyberAI);each category has its own scale.Sources:PitchBook Data,Inc.and SVB analysis.THE RISE OF CYBERAI|Tools and platforms that leverage

37、 GenAI to address vulnerabilities in AI infrastructure,models and applications or to defend against external,GenAI-driven threats and attacks.Including security operations,network security,identity&access management,endpoint security,data security and application security.Including model development

38、,vertical applications and computing infrastructure.Abnormal Securityis a developer of a cloud email security platform designed to prevent cybercrimes using an AI-based approach.Sandbox AQ is a developer of AI quantum software including an enterprise cryptography management solution.Pantheris an inf

39、rastructure security and compliance platform that uses AI to detect and respond to breaches.CertiKis a developer of a formal verification technology designed to protect and monitor blockchain protocols and smart contracts.Socure is a digital identity verification platform.The companys predictive ana

40、lytics platform applies AI techniques to verify identity and authenticity.9THE RISE OF CYBERAI|Funds focused on AI and cybersecurity have been a bright spot in an otherwise dull VC fundraising environment.These funds accounted for 17%of all 2023 vintage US VC funds.In some sense,this is not surprisi

41、ng.Among funds that raised in 2023,those mentioning AI as a specialty were nearly twice as likely to close their funds as those that didnt.1These funds have a broad base of companies to pursue,and investment into AI companies has skyrocketed over the past several years.Between 2020 and 2023,the numb

42、er of CyberAI companies formed was nearly 2x the number formed in the previous four-year period.Deal activity has kept pace,at least at the early stage.Besides the well-known VCs,corporates have funded much of the innovation and acquisition in CyberAI.Palo Alto Networks,for instance,has completed mo

43、re than 15 acquisitions since 2018 when Nikesh Arora took over as CEO.This acquisition streak is part of Aroras strategy to develop a cybersecurity platform,especially in the age of AI,noting in a 2023 interview that GenAI“should definitely create more opportunitynot just for security but for tech c

44、ompanies.”The corporate focus on consolidation is not unique to Palo Alto Networks,with other notable acquisitions such as CrowdStrikes March 2024 purchase of Flow Security,a cloud data runtime security solution,to bolster the parents product offerings.104875819911411820182019202020212022202355%VC21

45、%Angels11%CVC3%Accel/Inc.4%Private Equity5%Other051015202530354045Pre-201020102011201220132014201520162017201820192020202120222023THE RISE OF CYBERAI|As a%of All VC Funds RaisedAI Only Funds RaisedNotes:1)US-based VC funds with a stated sector focus of either AI or cybersecurity.“AI/Cyber”refers to

46、funds that had a stated focus in both categories.2)Company formation year determined using the founding date according to PitchBook Data,Inc.or its first financing date if the founding year is unavailable.3)Stage defined using PitchBook Data,Inc.s definition.Late stage defined as deals done five yea

47、rs after a companys founding.4)Only includes VCs that publicly disclosed investments.CyberAI VCs are aggregated from PitchBook Data,Inc.and Momentum Cyber data.Sources:Microsoft,CrowdStrike,Momentum Cybers H1 2023 Cybersecurity Market Review,The Channel Co.CRN,Preqin,PitchBook Data,Inc.and SVB analy

48、sis.Deal CountLate Stage3 VC InvestmentEarly Stage3 VC InvestmentSeed Investment0%5%10%15%20%0100200300400201820192020202120222023%of Total US VC Funds RaisedFund CountVintage YearCybersecurity Only Funds RaisedAI/Cyber Funds Raised$0.27B$0.56B$0.77B$2.05B$2.38B$1.02BIn 2022,VCs began issuing instru

49、ctions to portfolio companies to extend runway,cut burn and focus on efficiency over growth,typified by a16zs“A Framework for Navigating Down Markets”posted in May 2022.Since then,founders have largely followed their advice.Beginning in 2022,revenue growth began trending down for VC-backed tech comp

50、anies.At the end of 2021,annual revenue growth hovered around 40%-50%.By the end of 2023,it was less than half that.Companies are increasingly focused on profitability,often by shrinking expenses or cutting growth initiatives.This strategy is reflected in lower EBITDA margins as well.For AI and cybe

51、rsecurity companies,however,it is a different story.These companies continue to have burn that often far exceeds that of other VC-backed tech companies,and AI companies have actually increased their runway over the past three years,thanks to a favorable investment environment.Still,companies in the

52、space are dealing with the same declining revenue growth seen in other sectors.Part of this is defining a new sales process.While cybersecurity firms typically sell to a CISO,CyberAI companies have a less certain sales path.Many companies today are selling into chief risk officers(CROs),but going fo

53、rward we might expect to see the establishment of more AI committees to lead the acquisition of CyberAI solutions at an enterprise level.11Notes:1)Sector definitions based on SVB proprietary taxonomy.Data only for US companies with more than$5M in revenue.2)As of Q4 of each year.Sources:SVB propriet

54、ary data and SVB analysis.THE RISE OF CYBERAI|OtherAICybersecurity OtherAICybersecurity 0 Mos.5 Mos.10 Mos.15 Mos.20 Mos.25 Mos.201820192020202120222023201820192020202120222023201820192020202120222023OtherAICybersecurityOtherAICybersecurity-120%-100%-80%-60%-40%-20%0%2018Q42019Q42020Q42021Q42022Q420

55、23Q40%10%20%30%40%50%60%70%80%2018Q42019Q22019Q42020Q22020Q42021Q22021Q42022Q22022Q42023Q22023Q40.0 x0.5x1.0 x1.5x2.0 x2.5x3.0 x2018Q42019Q42020Q42021Q42022Q42023Q412THE RISE OF CYBERAI|0%32%58%11%Not a PriorityNot a Major PriorityTop 5 PrioritiesMost Critical PriorityDeveloping Capabilities In-Hous

56、eLeveraging Existing VendorsAdopting New GenAI SolutionsVery Important42%21%21%Somewhat Important37%53%37%Neutral11%16%37%Somewhat Unimportant11%5%5%Very Unimportant0%5%0%47%26%16%11%42%37%21%Security Built Into the SolutionItselfExisting CybersecurityPlatforms/ToolsNew CybersecurityPlatforms/ToolsT

57、here is an urgency to leverage GenAI within organizations,according to a panel of CISOs polled for this report.1 More than two-thirds of CISOs report that GenAI is among the top five priorities at their organizations,putting increased pressure on CISOs to strategically manage the rollout.To manage i

58、mplementation and the security concerns associated with these new GenAI tools,most CISOs are focused on developing security capabilities in-house.At a high level,this is to maintain control and minimize external risks.Still,integrating external solutions can be significantly easier than developing n

59、ew capabilities internally,and 74%of CISOs see this as important in the near term.Developing and deploying GenAI tools,however,is only half the battle;securing these systems while in use is critical.Fortunately,most CISOs either have a specific GenAI policy or are applying an existing general framew

60、ork successfully.CISOs often turn to traditional cybersecurity tools,with some seeing GenAI as more of a capability vs.a specific toolset to secure against.Others continue to evaluate their options,seeking tools that enable security of various aspects of LLMs,from training data to live usage.In term

61、s of specifics,CISOs report a wide range of tools used for GenAI security.This could reflect the industrys early stage of development,with more consolidation likely in the future.13THE RISE OF CYBERAI|In-HouseExistingNewVery Important42%32%26%Somewhat Important42%47%47%Neutral11%16%21%Somewhat Unimp

62、ortant0%5%0%Very Unimportant5%0%5%Defining a security policy is a gating item before we pilot GenAI tools.Our GenAI policy is well-defined,specific and deployed.Our GenAI-specific security policy is in development.We have a general security policy framework that we apply to AI successfully.Access Co

63、ntrolsCSPMIn-house Built PlatformAbnormal SecurityDLPMicrosoft CopilotAWSDSPMNetskopeCalypso AIGitHub CopilotNetwork FirewallsCASBIAMUser/Data MonitoringNew Cybersecurity ToolsExisting Cybersecurity ToolsSecurity Built into the Solution ItselfNotes:1)SVB survey fielded in April 2024 of a curated gro

64、up of CISOs with experience implementing enterprise GenAI cybersecurity solutions.2)Selected responses from survey respondents.Sources:SVB survey and SVB analysis.As enterprises deploy GenAI tools at scale,CISOs are reacting quickly to update their cybersecurity positioning and guard their organizat

65、ions from a wider attack surface.Just 18 months post-release of ChatGPT,74%of CISOs have already changed the composition of their cybersecurity stack,with the majority of that group primarily augmenting their existing stack with new tools.Among the 26%who have not yet changed their cybersecurity sta

66、ck,nearly all will add new cybersecurity tools within the next three years.This is a growth opportunity for CyberAI startups,which are well-poised to fill this gap.When surveyed,1 CISOs report a wide variety of AI tools that are impacting their organizations,and fortunately for CyberAI firms,CISOs e

67、xpect a 13%increase of cybersecurity budgets on average.This suggests that there is a wide opportunity set for CyberAI startups to address security issues around everything from code generation tools to knowledge management systems.One of the VCs interviewed for this report explained the opportunity

68、 for CyberAI solutions:“Attack campaigns that would have been carried out over months will be compressed to days or hours.We have very little defense against these intelligent attacks.Automation in our defenses is becoming more essential.The future will increasingly become AI versus AI.”14THE RISE O

69、F CYBERAI|26%47%21%5%We have replaced meaningful portions of our cybersecurity stack.We have primarily leveraged new capabilities from our existing cybersecurity stack.We have primarily augmented the capabilities of our existing stack with new tools.We have not meaningfully changed our cybersecurity

70、 stack.5%68%21%5%We expect to replace meaningful portions of our cybersecurity stack.We expect to primarily leverage new capabilities from our existing stack.We expect to primarily augment the capabilities of our existing stack with new tools.We do not expect to meaningfully change our cybersecurity

71、 stack.0%-10%-15%-5%+13%+15%+5%13%increase in budget expected on averageNotes:1)SVB survey fielded in April 2024 of a curated group of CISOs with experience implementing enterprise GenAI cybersecurity solutions.2)Selected responses from survey respondents.Sources:SVB survey and SVB analysis.15THE RI

72、SE OF CYBERAI|M&A 82%IPO 4%No Exit 14%CyberAI is an emerging space,still too early in the development cycle for many significant exits.But as investors and founders approach the industry,how are they thinking about future exit opportunities?Early in the development of a new vertical,M&A tends to be

73、the most common exit route.CyberAI is proving to be no different.Among recent large deals,Gem Security,a cloud detection and response firm,was bought by Wiz for$350M,and Silk Security,a cyber risk platform,was bought by Armis for$150M.Strong performance by related public companies and significant ca

74、sh on hand provide strong tailwinds for additional acquisition activity.IPOs could take more time,however.One interviewee noted that“It is too early to tell if cybersecurity and AI categories can support large IPOs,but Id bet the early demand from customers will drive approximately$100M+in M&A.”Pitc

75、hBook Data,Inc.s exit expectation analysis for the CyberAI cohort suggests that while most have a high probability of a successful exit,few will be via IPO.Still,it could just be a matter of time,with one interviewee predicting that“For IPOs,the time horizon is more like 3-5 years or more.”The indus

76、try is in the early stages of growth and development,with one expert explaining,“AI will dramatically change cyber,but it will not happen overnight.We are on a 3-5 year journey,and we are barely one year into it.”16Notes:1)Public AI cohort is an equal weighted basket of public AI companies including

77、 META,MSFT,GOOGL,AMZN,NVDA and TSM.2)Public cyber cohort is an equal weighted basket of public cyber companies including CRWD,ZS,OKTA,S,NET,CYBR,PANW and FTNT.3)Exit probabilities for each company are calculated by PitchBook Data,Inc.s exit probability methodology.Sources:CTech,Armis,S&P Capital IQ,

78、PitchBook Data,Inc.and SVB analysis.S&P 500Renaissance Capital IPO ETFPublic Cyber Cohort2Spread:Renaissance to S&P 500MedianAverage 93%85%75%64%Successful Exit75th percentile25th percentileTHE RISE OF CYBERAI|-50%0%50%100%150%200%250%300%350%400%Jan 20Apr 20Jul 20Oct 20Jan 21Apr 21Jul 21Oct 21Jan 2

79、2Apr 22Jul 22Oct 22Jan 23Apr 23Jul 23Oct 23Jan 24Apr 24Public AI Cohort1$327M$407M$514M$465M$426M$557M15.7%17.9%14.4%12.6%13.1%15.7%201820192020202120222023S&P 500 Tech Cash as%of Total S&P 500 CashS&P 500 Tech CashSuccessfulExit“AI is going to raisethe table stakes for cybersecurity companies.Those

80、 that integrate the new functionality effectively will differentiate themselves,deliver more value to their customers and will be better received in the capital markets,be it M&A or IPO.”“Ecommerce on the internet started in 1996,but it was not until an industry reboot that in 2002 ecommerce took of

81、f as we know it today.CyberAI is still in 1996 or 1997.We are still in the experimentation or forming stage.”36%25%15%8%No Exit17Andrew McCartyVice PresidentTech BankingSilicon Valley BAndrew Pardo,CFASenior Analytics ResearcherSVB Market InsightsSilicon Valley BEmma EschweilerDirectorInvestor Cover

82、ageSilicon Valley BJake Ledbetter,CFASenior Analytics ResearcherSVB Market InsightsSilicon Valley BAndrew McCarty is a Senior Vice President on SVBs NorCal Enterprise Software team.In this role,he advises a portfolio of venture-backed enterprise software clients in the San Francisco Bay Area from se

83、ed to pre-IPO.Andrew advises startups and their founders on creative banking and financing solutions while leveraging SVBs network to maximize clients probability of success.Before his current role,Andrew spent six years on SVBs NorCal Credit Solutions team,specializing in structuring and deploying

84、debt facilities for enterprise software clients.Andrew graduated from Santa Clara University with a bachelors degree in history and political science and a minor in economics.Emma Eschweiler is a Director for SVBs Investor Coverage&Business Development team.She is responsible for the growth and deve

85、lopment of institutional VC and corporate relationships on behalf of SVB,with a focus on partners making direct investments,acquisitions or partnerships principally within the enterprise software ecosystem,including areas such as cybersecurity,developer tools,applications,data infrastructure and ope

86、n-source initiatives.Before her current role,Emma was a vice president in SVBs NorCal Enterprise Practice,where she managed a portfolio of growth-stage,venture-backed enterprise software clients in the region.Emma earned her bachelors degree in government and global studies from Colby College and co

87、ntinued her education at Georgetown University and Sciences Po in France.Natalie Fratto Managing DirectorInvestor CoverageSilicon Valley BNatalie Fratto is a Managing Director on SVBs Investor Coverage&Business Development team.In this role,she seeks to be a thought partner to general partners on th

88、e financial journeys of both their portfolio companies and their firms.Earlier in her finance career,Natalie opened the SVB Canada office and led alternative investing into venture funds as part of the Corporate Development&Strategy team at Goldman Sachs.Before entering finance,she led operations fo

89、r a Y Combinator-backed services marketplace company and spent time as a management consultant focused on M&A within the semiconductor industry.Ever passionate about new technologies,Natalie writes and contributes technology pieces for Fortune,Motherboard and other tech outlets.Her viral 2019 TED Ta

90、lk on adaptability has been viewed four million times and was one of the top 10 most-watched talks of the year.THE RISE OF CYBERAI|Information for this report was gathered from a variety of sources.Quantitative data was sourced from proprietary SVB databases,industry publications and data providers

91、such as PitchBook Data,Inc.Quotes and many of the insights included in the commentaries are sourced from conversations with investors and operators in the CyberAI space,all of which were conducted during April and May 2024.Survey respondents were recruited from a curated list of CISOs who are SVB cl

92、ients.Each of the CISOs has experience leading organizational adoption of new technologies and brings a wealth of experience in the CyberAI space.All respondents are US-based.Silicon Valley Bank(SVB),a division of First-Citizens Bank,is the bank of some of the worlds most innovative companies and in

93、vestors.SVB provides commercial and private banking to individuals and companies in the technology,life science and healthcare,private equity,venture capital and premium wine industries.SVB operates in centers of innovation throughout the United States,serving the unique needs of its dynamic clients

94、 with deep sector expertise,insights and connections.SVBs parent company,First Citizens BancShares,Inc.(NASDAQ:FCNCA),is a top 20 U.S.financial institution with more than$200 billion in assets.First Citizens Bank,Member FDIC.Learn more at .Silicon Valley B See complete disclaimers on the following p

95、age.2024 First-Citizens Bank&Trust Company.Silicon Valley Bank,a division of First-Citizens Bank&Trust Company.Member FDIC.3003 Tasman Drive,Santa Clara,CA 95054 1818THE RISE OF CYBERAI|The views expressed in this report are solely those of the authors and do not necessarily reflect the views of SVB

96、.This material,including without limitation to the statistical information herein,is provided for informational purposes only.The material is based in part on information from third-party sources that we believe to be reliable but which has not been independently verified by us,and,as such,we do not

97、 represent the information is accurate or complete.The information should not be viewed as tax,accounting,investment,legal or other advice,nor is it to be relied on in making an investment or other decision.You should obtain relevant and specific professional advice before making any investment deci

98、sion.Nothing relating to the material should be construed as a solicitation,offer or recommendation to acquire or dispose of any investment,or to engage in any other transaction.All non-SVB named companies listed throughout this document,as represented with the various statistical,thoughts,analysis

99、and insights shared in this document,are independent third parties and are not affiliated with Silicon Valley Bank,a division of First-Citizens Bank&Trust Company.Any predictions are based on subjective assessments and assumptions.Accordingly,any predictions,projections or analysis should not be vie

100、wed as factual and should not be relied upon as an accurate prediction of future results.Investment Products:Are not insured by the FDIC or any other federal government agencyAre not deposits of or guaranteed by a bankMay lose value2024 First-Citizens Bank&Trust Company.Silicon Valley Bank,a division of First-Citizens Bank&Trust Company.Member FDIC.3003 Tasman Drive,Santa Clara,CA 950541919THE RISE OF CYBERAI|